Skip to Content

Finding our voices: Capgemini’s blueprint to advance women and drive success in A&D

Elodie Regis
Elodie Regis
Jul 23, 2024

In the male-dominated corridors of the Aerospace and Defense (A&D) industry, I have sometimes felt different. Yet, it was this difference that ignited my mission to create a community where women could find their voices and support each other.

In January 2024, with the support of Capgemini’s leadership, I started a Women in A&D Community at Capgemini, a part of the global Women @ Capgemini Employee Network Group. This initiative was born out of a desire to create a platform where women could connect, mentor, and inspire one another. Our A&D leadership wanted to mobilize women and men around their collective interest in and passion for the A&D industry. We wanted to create a strong internal network in an effort to attract more women into client service in A&D.

However, we also saw the opportunity to make our Capgemini women more visible externally by sharing their stories and demonstrating how women can succeed in A&D. The impact was intended to be twofold: sharing our own story with clients so we can share best practices and positioning Capgemini as a place for young women in A&D to build careers.

This initiative isn’t just about supporting women. It’s also about driving business success. One study from the Peterson Institute for International Economics showed that companies with greater gender diversity at the C-suite level are more profitable. Women bring unique perspectives and leadership styles that enhance innovation and decision-making. By fostering an environment where women can thrive, A&D organizations can tap into a broader talent pool and achieve long-term success in a competitive landscape.

With collaboration as the theme for the upcoming Farnborough Air Show, I saw this as the perfect opportunity to share our journey, the lessons we’ve learned, and how other organizations can benefit from establishing similar network groups. Through collaboration, we can create inclusive futures for everyone in our industry.

Building a community of support and empowerment

I vividly recall nights spent working on presentations after putting my son to bed, feeling exhausted but determined. I’ve heard similar stories from other women—of juggling work and personal life, of feeling sidelined in meetings, and of striving to be heard and seen. The Women in A&D network group quickly became a vibrant community, providing the much-needed support and mentorship that I have longed for.

The A&D industry has long been a challenging environment for women. Stories shared within our network group and broader industry experiences reveal the myriad obstacles women face in this predominantly male field. The scarcity of female role models and limited support networks can leave women feeling isolated, while microaggressions and subtle biases chip away at their confidence.

The struggle to balance demanding careers with personal responsibilities adds another layer of complexity, and women often find their voices overshadowed or underestimated. These systemic issues underscore the urgent need for dedicated support and advocacy.

At Capgemini, our Women in A&D community tackled these challenges head-on by fostering open and honest discussions in a safe space. One of our key initiatives was a series of monthly webinars featuring female industry leaders sharing their journeys and the obstacles they overcame. These stories, like one from a speaker with over three decades of A&D experience who recounted being ignored on several occasions in meetings, left a lasting impact on our members.

We also organized coaching sessions focused on developing essential professional skills, from pitching oneself effectively to articulating career goals. These sessions empowered women to take control of their careers, like a young manager who found the courage to pursue a director position while planning to start a family.

Mentorship plays a crucial role, with experienced women guiding their less experienced peers. As a mentor, I have found great satisfaction in sharing my own journey and providing practical advice. The connections forged through these relationships have been transformative, and mutually beneficial.

Throughout our discussions, the need for visible female role models emerged as a recurring theme. We celebrated successes, no matter how small, to inspire others. One member’s promotion to a senior engineering role, which she attributed to the confidence and skills gained from our community, served as a powerful reminder of the impact we can have when we support each other.

What we’ve learned from the women in our community

Supporting women in the male-dominated aerospace and defense industry is no easy feat, but through our Women in A&D community at Capgemini, we have gained invaluable insights that light the way forward:

Support networks can make all the difference. In an industry where women often feel isolated, having a dedicated community that understands their unique challenges provides solace and strength. Imagine the boost in confidence when you know you’re not alone in navigating difficult projects or workplace biases.

Mentorship is power. Nothing accelerates growth quite like guidance from those who have walked the path before. When experienced professionals share their stories of overcoming gender biases and balancing work with personal life, it creates a ripple effect. Our network group’s mentors have guided women to pursue ambitious goals without compromising their wellbeing.

Work-life balance remains an uphill battle, one that demands organizational support. Too many women find themselves torn between demanding careers and personal responsibilities. By offering flexible solutions and empathetic policies, we can help them thrive in both realms.

Visibility is vital. Women’s contributions must be seen and celebrated. By regularly spotlighting their achievements, we inspire others and foster a culture of recognition. We share each success story in our community, whether a hard-fought promotion or a skillful presentation.

Open discussions drive change. Only by having honest conversations about biases and challenges can we raise awareness and shift culture. When a senior female leader shares her experience of being ignored in meetings and it sparks a wave of similar stories, the collective dialogue becomes a force for progress.

Skill-sharpening builds confidence. Coaching sessions on pitching, articulating career goals, and building confidence are central for professional growth. With these tools in their arsenal, women are empowered to advocate for themselves and seize every opportunity.

Men are welcome! It’s not just women helping women, we need men at the table with women to advocate for them. Male mentorship is just as important as female mentorship as it fosters collaboration and sharing differences of personal experiences and skill sets. Men in leadership positions are in the position to promote their female colleagues and position them in leadership roles thus creating a culture of female advocacy and advancement. Men “pounding the table” for women reinforces confidence in the valuable perspectives and insights that women bring to the table.

Putting our insights into action: a guide for A&D organizations

This is about more than just meeting the needs of female employees – it’s a strategic business imperative.

By creating an environment where women can thrive, A&D organizations tap into a deep pool of talent and position themselves for long-term success in an increasingly complex and competitive landscape.

Embracing strategies to support and advance women isn’t just about checking a box or making incremental improvements. It’s about fundamentally reshaping the culture and priorities of the A&D industry to harness the full spectrum of human talent and insight.

Here’s how organizations can do that:

  1. Build vibrant support networks. Create communities where women can connect, share, and support each other through regular meetups, online forums, and social media groups. Give them a safe haven to express challenges and find solutions together.
  2. Make mentorship a priority. Encourage experienced professionals to guide and uplift those starting out. Formalize mentoring programs with structured sessions, goals, and progress tracking to magnify the impact.
  3. Champion work-life balance. Show you value your female talent by offering flexible working arrangements and support systems that accommodate the dual pressures they face. When women can balance career and personal life, everyone wins.
  4. Put achievements in the spotlight. Regularly highlight and recognize the successes of women across your organization through newsletters, award ceremonies, and internal communications. Make it clear that their contributions are seen and valued.
  5. Start conversations that matter. Foster an inclusive environment by encouraging honest discussions about challenges and biases. Create platforms for dialogue like town halls, panels, and anonymous feedback channels. Each conversation is a catalyst for change.
  6. Invest in their growth. Provide coaching and training to help women sharpen their professional skills and confidence. Offer workshops on public speaking, leadership, negotiation, and career planning. Empower them to advocate for themselves and pursue their ambitions fearlessly.
  7. Men are a critical part of the community. This is about inclusivity, not exclusivity. We welcome anyone and everyone to the conversation, not just women. Men are an important part of advancing women as it demonstrates how expanding a team to a difference in opinions, experiences, and ways of thinking can enhance project management and delivery. 

By weaving these practices into the fabric of your organization, you can create an environment where women in A&D don’t just survive but thrive every day.

Learn more:

Digital Continuity in the Aerospace Industry

Digital Twins in Aerospace and Defense

Intelligent Supply Chain for the Aerospace and Defense Industry

Lifecycle Optimization for Aerospace and Defense

 

Mee the author

Elodie Regis

Elodie Regis

VP, Aerospace & Defense, Capgemini Invent
Elodie is Vice President at Capgemini Invent, leading 2 main topics : industrial ramp-up in the Aerospace & Defense and Skywise. She has a diverse background including Quality Director in a factory for the Automotive Industry and work as a Consultant for 18 years. She has developed her expertise on A&D Manufacturing, Quality and Supply Chain while designing and building new factories, supporting shopfloor workforce transformation, and operations excellence.

    Small is the new big: The rise of small language models

    Sunita Tiwary
    Jul 22, 2024

    In the dynamic realm of artificial intelligence (AI) and machine learning, a compelling shift is taking center stage: the ascent of small language models (SLMs). The tech world is smitten with the race to build and use large, complex models boasting billions and trillions of parameters, and consumers have become unwitting accomplices in the obsession with “large”. However, recent trends indicate a growing interest in smaller, more efficient models. This article delves into the reasons behind this shift, its implications, and what it means for the future of AI.

    Before we dive into SLMs, how did the wave of large languages grow

    In the not-so-distant past, natural language processing (NLP) was deemed too intricate and nuanced for modern AI. Then, in November 2022, OpenAI introduced ChatGPT, and within a mere week, it garnered more than a million users. Suddenly, AI, once confined to research and academic circles, became accessible to the masses. For example, my nine-year-old daughter effortlessly began using ChatGPT for school research tasks, while my mother-in-law, in her late sixties, whose only tech acquaintance was limited to WhatsApp and Facebook, now enthusiastically shares the latest news about AI, and her budding interest in GenAI during our tea time conversations.

    The launch of ChatGPT marked the onset of the very loud and very public (and costly) GenAI revolution, effectively democratizing AI. This is evident in integrating AI as copilots in various products, the exponential growth of large language models (LLMs), and the rise of numerous startups in this space. The landscape of technology and our world will never be the same.

    To comprehend the magnitude of this shift, let’s delve into the parameters of AI models. The number of parameters is a core measure of an AI’s scale and complexity. GPT-2 had 1.5 billion parameters, and then OpenAI released GPT-3, which had a whopping 175 billion parameters. This was the largest neural network ever created, more than a hundred times larger than its predecessor just a year earlier. Now we see a trillion parameter LLMs.

    Deciphering SLMs

    While the definition of an SLM remains contextual, some research identifies them as models encompassing approximately 10 billion parameters or less. SLMs are lightweight neural networks that can process natural language with fewer parameters and computational resources than LLMs. Unlike LLMs (which are generalized models), SLMs are usually purpose-driven and tailored to address specific tasks, applications, or use cases.

    Recent studies demonstrate that SLMs can be fine-tuned to achieve comparable or even superior performance compared to their larger counterparts in specific tasks.

    For example, phi-3-mini, a 3.8 billion parameter SLM trained on 3.3 trillion tokens, whose overall performance, as measured by both academic benchmarks and internal testing, rivals that of models such as Mixtral 8x7B and GPT-3.5 (e.g., phi-3-mini achieves 69 percent on MMLU and 8.38 on MT-bench). Another example is phi-3-vision, a 4.2 billion parameter model based on phi-3-mini with strong reasoning capabilities for image and text prompts. Similarly phi-2 matches or outperforms models up to 25 times larger on complex benchmarks. Another such model is Orca2, which was built for research purposes. Similarly, TinyLlama, launched in late 2023, had just 1B parameters, followed by the OpenELM by Apple for edge devices launched in April 2024.

    Why does it matter?

    SLMs bring many benefits, notably their swift training and faster inference speed. Beyond efficiency, these models contribute to a more sustainable footprint, showcasing reduced carbon and water usage. In addition, SLMs strike a harmonious balance between performance and resource efficiency. Training SLMs is much more cost-effective due to the reduced number of parameters and offloading the processing workload to edge devices further decreases infrastructure and operating costs.

    Credit: Microsoft

    1. Efficiency and sustainability

    It is crucial to acknowledge that LLMs demand substantial computational resources and energy. Complex architecture and vast parameters necessitate significant processing power that contributes to environmental and sustainability concerns.

    In contrast, SLMs significantly reduce computational and power consumption through several key factors:

    • Reduced computational load: Small models have fewer parameters and require less computation during inference, leading to lower power consumption
    • Shorter processing time: The reduced model size decreases the time required to process inputs thus consuming less energy per task
    • Lower memory usage: Smaller models need less memory, which reduces the power needed for memory access and management which is a significant factor in energy consumption. Efficient use of memory further minimizes the energy needed to store and retrieve parameters and intermediate calculations
    • Thermal management: Lower computational requirements generate less heat, reducing the need for power-hungry cooling systems. Furthermore, reduced thermal stress increases the longevity of hardware components, indirectly reducing the energy and resources needed to replace and maintain them.

    SLMs are increasingly becoming popular due to their efficiency. They require less computational resources and storage than LLMs, making them a more practical solution for many applications requiring real-time processing or deployment on edge devices with limited resources. Therefore, by reducing the model size and complexity, developers can achieve faster inference times, lower latency, and improved performance, making small models preferred for resource-constrained environments such as mobile phones, personal computers, or connected devices. For example, phi-3 is highly capable of running locally on a cell phone. Phi-3 can be quantized to four bits so it occupies only ~1.8GB of memory. The quantized model of phi-3 when tested on iPhone 14 with A16 Bionic chip running natively on-device and fully offline achieving more than 12 tokens per second (the rate at which a model processes tokens (words, subwords, or characters) during inference).

    According to the Tirias Research GenAI Forecast and TCO Model, if 20 percent of GenAI processing workload could be offloaded from data centers by 2028 using on-device and hybrid processing, then the cost of data center infrastructure and operating cost for GenAI processing would decline by $15 billion (where data center infrastructure and operating costs projected to increase to over $76 billion by 2028.). This also reduces the overall data center power requirements for GenAI applications by 800 megawatts.

    2. Economic viability

    Developing and maintaining LLMs comes with steep costs, demanding significant investments in computational resources, energy usage, and specialized skills. In contrast, SLMs present a more budget-friendly solution. Their streamlined design means they are more efficient at training and require less data and hardware, leading to more economical computing costs. SLMs often employ optimized algorithms and architectures designed for efficiency. Techniques like pruning (removing unnecessary parameters) and quantization (using lower precision arithmetic) make these more economically viable.

    3. Scalability and accessibility

    Smaller models are inherently more scalable and accessible than their larger counterparts. By reducing model size and complexity, developers can deploy AI applications across various devices and platforms, including smartphones, IoT devices, and embedded systems. This democratizes AI, encourages wider adoption, and accelerates innovation, unlocking new opportunities across many industries and use cases.

    4. Ethical and regulatory dimensions

    Ethical and regulatory considerations also contribute to the shift towards SLMs. As AI technologies become increasingly pervasive, data privacy, security, and bias concerns become more pronounced. Embracing small models allows organizations to reduce data exposure, address privacy challenges, and reinforce transparency and accountability. When trained on specific, high-quality datasets, smaller models significantly reduce the risk of data exposure. They require less training data compared to their larger counterparts, which lowers the risk of memorizing, overfitting, and inadvertently revealing sensitive information within the training set. With fewer parameters, these models have simpler architectures, minimizing potential pathways for data leakage. Furthermore, smaller models are easier to interpret, validate, and regulate, facilitating compliance with emerging regulatory frameworks and ethical guidelines.

    Limitations of SLMs

    While SLMs have great benefits, there are challenges and limitations too. Due to their smaller size, these models do not have the capacity to store too much “factual knowledge” This could lead to hallucination, factual inaccuracies, amplification of biases, inappropriate content generation, and safety issues. However, this can be mitigated by the use of carefully curated training data and targeted post-training and improvements from red teaming insight. Models can also be augmented with a search engine for factual knowledge.

    Conclusion

    The transition to SLMs represents a significant trend in the AI field. While LLMs excel due to their vast size, intensive training, and advanced NLP capabilities, SLMs offer targeted efficiency, cost-effectiveness, and scalability. By adopting these models, organizations can unlock new opportunities, speed up innovation, and create value across various sectors.

    The future of generative AI is also moving towards the edge, enabled by small, efficient language models. These models transform everyday technology with natural, generative interfaces, encompassing everything from personal devices and home automation to industrial machinery and intelligent cities.

    SLMs are essential to enable AI at the edge. According to IBM, Huawei, and Grand View Research, the edge AI market is valued at $21 billion and is expected to grow at a CAGR of 21 percent. Companies like Google, Samsung, and Microsoft are advancing generative AI for PCs, mobile, and connected devices. Apple is joining this effort with OpenELM, a group of open-source LLMs and SLMs designed to run entirely on a single device without cloud server connections. This model, optimized for on-device use, can handle AI tasks independently, marking a new era in mobile AI innovation, as noted by Alphasense.

    Finally, it’s not a matter of choosing one over the other. LLMs are generalists with extensive training in massive data and have extensive knowledge across various subjects. They have the ability to perform complex interactions like chatbots, content summarization, and information retrieval and have vast applicability, however, are expensive and have a high operational cost. SLMs on the other hand are specialized, domain-specific powerful, and less computationally intensive but struggle with complex context and hallucinations if not used on their specific use case and context. The choice between SLM and LLM is dependent on the need and availability of resources, nevertheless, SLM is surely a game changer in the AI era.

    Author

    Sunita Tiwary

    Senior Director– Global Tech & Digital
    Sunita Tiwary is the GenAI Priority leader at Capgemini for Tech & Digital Industry. A thought leader who comes with a strategic perspective to Gen AI and Industry knowledge. She comes with close to 20 years of diverse experience across strategic partnership, business development, presales, and delivery. In her previous role in Microsoft, she was leading one of the strategic partnerships and co-creating solutions to accelerate market growth in the India SMB segment. She is an engineer with technical certifications across Data & AI, Cloud & CRM. In addition, she has a strong commitment to promoting Diversity and Inclusion and championed key initiatives during her tenure at Microsoft.

    Fabio Fusco​

    Data & AI for Connected Products Centre of Excellence Director​, Hybrid Intelligence​, Capgemini Engineering
    Fabio brings over 20 years of extensive experience, blending cutting-edge technologies, data analytics, artificial intelligence, and deep domain expertise to tackle complex challenges in R&D and Engineering for diverse clients and is continuously forward-thinking.

      Leveraging digital twins to transform insurance operations

      Amit Bhaskar, Head of Financial Services, Capgemini’s Business Services
      Amit Bhaskar
      Jul 18, 2024

      Digital twins can deliver improvement recommendations and best practices to help insurers drive enhanced business outcomes from their insurance operations.

      The concept of digital twins can be defined as “a virtual representation of real-world entities and processes synchronized at a specified frequency and fidelity.”

      What is perhaps less obvious are the potential benefits digital twins can bring to the information processing domains of finance, HR, and supply chain management, where they can be used to avoid bottlenecks through problem prediction, increasing efficiency, and reducing downtime.

      Digital twins aim to bridge the “physical-digital” gap to improve performance and sustainability, establishing a closed-loop approach to unlock value through bringing synergies across data, technologies, and business processes. Digital twins not only provide visibility into how dynamic, real-world systems are performing, but can also predict how they will perform in different scenarios.

      Digital twins can augment insurance processing

      Insurers are investing in digital twin technologies to shift from being reactive to proactive and improve decision-making. Implementing digital twins can improve business areas including underwriting, claims processing, risk management, finance and accounting, and customer service functions.

      The digital twin is the foundation for data-driven transformation and continuous innovation.
      The digital twin is the foundation for data-driven transformation and continuous innovation
      • 40% of companies have an ongoing comprehensive digital twin program across the whole value chain, according to Gartner
      • 47% of organizations are considering strategic partnerships for digital twin initiatives, according to Capgemini Research Institute
      • 38% of organizations are willing to work with new/existing partners towards digital twin implementation, according to the same Capgemini report
      • The digital twin market will reach $183 billion of revenue by 2031, according to Gartner.

      Adding value to insurance operations

      In the insurance industry, digital twins could help insurers to increase process efficiency and operational effectiveness in areas that include:

      • Analyzing claims assessment and compensation lead times under different operating conditions to check for possible non-compliance with regulations
      • Simulating damage conditions and reproducing loss events for claims, enabling more accurate loss estimation and fraud detection
      • Identifying strengths and weaknesses, including capacity limitations, bottlenecks, critical point of failure, and interdependencies
      • Providing data and insights to support decision-making
      • Measuring the impact of potential changes prior to implementation in a live environment
      • Highlighting opportunities for continuous improvement
      • Creating and testing alternative flows and outcomes to decide how to deliver optimum value
      • Measuring real change outcomes with continuous business mining
      • Monitoring return on investment.

      Opportunities to drive business outcomes for insurance

      Leveraging digital twins can help insurers identify and implement improvement recommendations and best practices that deliver enhanced business outcomes from their insurance operations. This not only includes transformation roadmap design, but opportunities to improve visibility, efficiency, and working capital.

      Whether it’s streamlining the underwriting process, reshaping claims processing, enhancing customer service, and improving regulatory compliance, integrating digital twins into insurance can drive increased efficiency, personalization and customer engagement, boosting the value insurers deliver to their customers.

      To learn more about how Capgemini can help your insurance company leverage digital twins for data-driven transformation and continuous innovation across your insurance operations, contact: bhaskar.amit@capgemini.com or aneta.szporak@capgemini.com.

      Meet our experts

      Amit Bhaskar, Head of Financial Services, Capgemini’s Business Services

      Amit Bhaskar

      Head of Financial Services, Capgemini’s Business Services
      Amit Bhaskar helps our banking, capital markets, and insurance clients to transform, profit, and grow – leveraging the Frictionless Enterprise to change the way you think, the way you work, and the way you engage with customers and your value network.
      Aneta Szporak Global Insurance Offer Lead, Capgemini’s Business Services

      Aneta Szporak

      Global Insurance Offer Lead, Capgemini Business Services
      Aneta Szporak has extensive experience in the insurance industry, especially in operations, customer service, organizational management, and product development. She leads the insurance offer for Capgemini’s Business Services Global Business Line.

        Unlocking the power of data with speech analytics

        Amit Bhaskar, Head of Financial Services, Capgemini’s Business Services
        Amit Bhaskar
        Jul 18, 2024

        Implementing a next-generation contact center solution can help insurers drive meaningful outcomes from client interaction data to deliver more meaningful, emotive, and connected relationships with their policyholders.

        39% of insurance executives said they face hurdles in technology readiness on the journey to life insurance modernization, according to Capgemini’s Insurance Top Trends 2024. At the same time, policyholders face barriers related to product adoption, such as product complexity (39%), limited awareness (39%), and lack of trust (28%).

        According to Capgemini’s World Life Insurance Report 2023, 33% of the world’s population will be aged over 50 by 2050. To overcome these barriers and retain relevancy and trust among older policyholders, insurers can leverage data from multiple sources to generate single views of their customers. They can also utilize advanced technologies such as cloud computing, advanced analytics, and AI to digitally empower their agents and interact with customers.

        Speech analytics is one such technology for unraveling the potential of data to make informed decisions. But how can the analysis of speech drive business outcomes? And why does it hold such huge potential for the insurance industry?

        Driving business outcomes from interactions data

        Deriving business outcomes from sound is as much an art as it is a science. While harnessing cutting-edge speech analytics technology to unlock data is difficult enough, turning this data into meaningful business process outcomes represents a huge challenge for insurers.

        Indeed, organizations often struggle to apply data purposefully. But when it’s built around analyzing recorded calls to gather information on the persona, behavior, and needs of the individual customer, extracting meaningful insights from data buried in client interactions can enrich customer experience and drive operational efficiency.

        Leveraging AI-enabled speech analytics

        Leveraging Generative AI in contact center interactions can help insurers enhance operational efficiency and nurture strong customer relationships. Speech analytics leverages conversational intelligence that uses natural language processing and machine learning to convert spoken words into text for gaining insights into customer sentiments and needs.

        GenAI offers highly personalized interactions, superior customer experience, and heightened policyholder engagement and satisfaction – all while reducing operational costs. However, insurers must also grapple with the dual challenge of having to explain decisions based on GenAI, while also adhering to regulatory compliance in handling customer data and data privacy due to the sensitive nature of information within the insurance sector.

        In short, implementing a next-generation contact center solution that leverages a persona-influenced service design, AI, and advanced analytics can help you drive more meaningful, emotive, and connected relationships with your policyholders.

        To learn more about how Capgemini can help your insurance company unlock the true potential of data to drive a seamless, intelligent, and connected customer experience with your policyholders, contact: bhaskar.amit@capgemini.com or aneta.szporak@capgemini.com.

        Meet our experts

        Amit Bhaskar, Head of Financial Services, Capgemini’s Business Services

        Amit Bhaskar

        Head of Financial Services, Capgemini’s Business Services
        Amit Bhaskar helps our banking, capital markets, and insurance clients to transform, profit, and grow – leveraging the Frictionless Enterprise to change the way you think, the way you work, and the way you engage with customers and your value network.
        Aneta Szporak Global Insurance Offer Lead, Capgemini’s Business Services

        Aneta Szporak

        Global Insurance Offer Lead, Capgemini Business Services
        Aneta Szporak has extensive experience in the insurance industry, especially in operations, customer service, organizational management, and product development. She leads the insurance offer for Capgemini’s Business Services Global Business Line.

          Knowledge Graphs improve Gen AI
          Validating results builds trust for organizations

          Joakim Nilsson
          18th July 2024

          Generative AI can make recommendations that will transform decision-making for organizations – but how can people trust the answers Gen AI provides? Knowledge Graphs can play a vital role in ensuring the accuracy of Gen AI’s output, bolstering its reliability and effectiveness.

          In Douglas Adams’ The Hitchhiker’s Guide to the Galaxy, a supercomputer called Deep Thought is asked for the answer to “Life, the universe, and everything.” After 7.5 million years, Deep Thought responds “42.” Representatives from the civilization that built Deep Thought immediately ask how it arrived at the answer, but the computer cannot tell them. When Adams wrote this scene in the 1970s, he was (arguably) making a joke – but today, many people find themselves in this situation when interacting with generative AI (Gen AI).

          Gen AI works by drawing upon millions of pieces of data – a volume that’s impossible for humans to effectively analyze. Businesses are excited by its potential to deliver valuable insights and make well-informed predictions – but if different Gen AI tools are asked the same question and give different answers, how could an organization decide which result is more correct? How would a person fact-check the responses?

          Addressing the shortcomings of unstructured, implicit data

          The challenge relates to the Large Language Models Gen AI relies upon. An LLM can contain massive amounts of data, but it’s commonly stored in an unstructured, implicit manner. This makes it difficult to investigate how a Gen AI tool arrived at its answer.

          Since the release of ChatGPT in late 2022, Neo4j and Capgemini have been working independently / collaborating to overcome this challenge by using Knowledge Graphs. These store complex, structured data and the relationships between them. Instead of relying solely on LLMs to directly generate database queries, our solution incorporates a high-level interface that allows the LLM to interact seamlessly with a Knowledge Graph via database query templates. These templates serve as structured frameworks, guiding the LLM to fill in specific parameters based on the user’s request. This simplifies the task for the LLM by abstracting away complex logic. (See Figure 1.) This separation of concerns ensures the LLM focuses on natural language understanding and generation, while the query templates handle the technical aspects of database interaction – improving the overall accuracy and efficiency of retrieval.

          In this example, the query template uses a vector search to locate relevant nodes within the Knowledge Graph that correspond to the entities present in the user’s question. This identifies the nodes relevant to the query, which are then used to retrieve neighborhoods or shortest paths around the nodes within the graph. This helps contextualize the retrieved information and provides a more comprehensive answer to the user’s query. More information about this specific query template is available in this blog post.

          Tailored templates

          Query templates can be tailored to discrete domains such as finding dependencies within supply chains or executing aggregation operations for business intelligence purposes, enabling organizations to address specific challenges. This more targeted approach best leverages the LLM’s capabilities to generate insights by ensuring they are not only relevant but deeply informed by the underlying data structures, helping enterprises to efficiently transform their raw data into actionable intelligence.

          That said, the complexity of business requirements often exceeds what a single query template can accommodate when an LLM interfaces with a Knowledge Graph. Therefore, it’s essential to embrace an adaptive approach, providing a rich assortment of query templates that can be selectively deployed to match specific business scenarios. Leveraging the LLM’s capability to invoke functions, Gen AI can dynamically select and employ multiple query templates based on the context of the user’s request or the specific task at hand. This results in a more nuanced and flexible interaction with the database, and significantly amplifies the LLM’s ability to solve intricate business intelligence and analytics problems. (See Figure 2.)

          This LLM-powered movie agent uses several tools, orchestrated through carefully designed query templates, to interact with the Knowledge Graph.

          • The information tool retrieves data about movies or individuals, ensuring the agent has access to the latest and most relevant information.
          • The recommendation tool provides movie recommendations based on user preferences and input.
          • The memory tool stores information about user preferences in the Knowledge Graph, allowing for a personalized experience over multiple interactions.

          More information on this movie agent project can be found on GitHub.

          “We expect Knowledge Graphs to help Large Language Models embrace iterative processes to improve their output.”

          Democratizing data and empowering business users

          The Knowledge Graph acts as a bridge, translating user intent into specific, actionable queries the LLM can execute with increased accuracy and reliability. By allowing any user – regardless of technical knowledge – to inspect how the LLM arrived at its answers, people can validate the information sources themselves. Benefits include:

          • Results that are explainable, repeatable, and transparent. This can enhance trust in Gen AI in everything from research and discovery in life sciences to digital twins in sectors such as manufacturing, aerospace, and telecommunications.
          • Better-informed and better trusted business decisions
          • Freed up time for experts such as prompt engineers to concentrate on tasks that require their specialized skills.

          As we look ahead, we expect Knowledge Graphs to help Large Language Models embrace iterative processes to improve their output. Our enthusiasm is shared by other experts in the field including Andrew Ng at DeepLearningAI, underscoring the widespread recognition of their transformative capabilities. As we help create the future, it’s clear the journey with these intelligent systems is only just beginning – and is moving much faster than Deep Thought ever did – so it’s critical that people are given the means to fact-check generative AI as it evolves.

          INNOVATION TAKEAWAYS:

          TRUST IS IMPORTANT – Knowledge Graphs can boost confidence in the output from Gen AI systems – making it easier for people and organizations to embrace them.

          TOOLS FOR THE TOOL – With Knowledge Graphs, Large Language Models can dynamically employ multiple query templates to match specific business scenarios, making interactions with Gen AI more nuanced.

          DEMOCRATIZING DATA – By making it easier for everyone in an organization to interact with generative AI, Knowledge Graphs can free up experts to focus on tasks that require their specific skills.

          Interesting read?

          Capgemini’s Innovation publication, Data-powered Innovation Review | Wave 8 features contribution from leading experts from Capgemini and esteemed partners like Dassault Systèmes, Neo4j, and The Open Group. Delve into a myriad of topics on the concept of virtual twins, climate tech, and a compelling update from our ‘Gen Garage’ Labs, highlighting how data fosters sustainability, diversity, and inclusivity. Embark on a voyage of innovation today. Find all previous Waves here.

          Author

          Joakim Nilsson

          Knowledge Graph Lead, Insights & Data, Client Partner Lead – Neo4j Europe, Capgemini 
          Joakim is part of both the Swedish and European CTO office where he drives the expansion of Knowledge Graphs forward. He is also client partner lead for Neo4j in Europe and has experience running Knowledge Graph projects as a consultant both for Capgemini and Neo4j, both in private and public sector – in Sweden and abroad.

          Tomaz Bratanic

          Senior GenAI Developer, Neo4j
          Tomaz Bratanic has extensive experience with graphs, machine learning, and generative AI. He has written an in-depth book about using graph algorithms in practical examples. Nowadays, he focuses on generative AI and LLMs by contributing to popular frameworks like LangChain and LlamaIndex and writing blog posts about LLM-based applications.

          Magnus Carlsson

          CTO, Insights & Data Sweden, Capgemini
          Magnus heads Insights & Data Nordics at Capgemini. He focuses on innovation with data and how to scale. Some of examples of current areas he and his team are working on are Scaling AI & Data Science, Data Estate Modernization including Data Mesh, Datalakehouse and Knowledgraphs, User Adoption & Change Management, Data Governance / DataOps, Next Generation Analytics, CFO Office and not the least – Sustainability Analytics & AI.

            Identifying the best business applications for neutral atom quantum computers

            Camille de Valk
            Jul 18, 2024

            New research from Capgemini’s Quantum Lab focuses on both industry challenges and quantum computing theories to define practical, scalable and cost-efficient applications for neutral atom quantum computers.

            Neutral atom quantum computers are pushing the boundaries of quantum computing, allowing for more complex and higher-quality calculations. Exploring their potential and the most impactful use of this emerging tech is multifaceted and requires an understanding of its technological capabilities, industry-specific expertise, and a mastery of the quantum computing theory. Capgemini’s recent investment in Pasqal, a pioneer in quantum computing, known for creating innovative quantum processors using neutral atoms arranged in 2D and 3D structures, demonstrates the growing interest in the application of quantum computing to deliver clear business value. How do we move from the abstract and theoretical, and how exactly do we evaluate what may or may not be a good application?    

            A good methodology to evaluate what good applications look like 

            To evaluate whether an application is suitable for neutral atom quantum computers, one must consider the end-to-end algorithm, considering both classical and quantum calculations. For example, neutral atom quantum computers, when running in analog mode, solve a specific mathematical problem. That is to say that it finds a maximum independent set (MIS) [see definitions] in a unit disk graph (UDG) [see definitions]. Finding an MIS in a UDG means identifying the largest possible group of nodes within the graph that are not directly connected to each other by an edge, see Figure 1. The class of combinatorial optimization problems, however, is a lot bigger than just the class of MIS problems on a UDG. This means that not all combinatorial optimization problems can be solved by the neutral atom quantum computer natively. The problems must be mapped to a form that can be solved by the neutral atom quantum computer. This can add significant computational costs, which cannot be neglected but sometimes are. The result of cutting these costs can drastically favor the quantum computer, which is unfair. That is why end-to-end considerations must be made.  

            Figure 1 an example of a non-independent set (turquoise), independent set (green), and maximum independent set (dark blue). The red nodes don’t form a independent set because there are nodes that share an edge. The turquoise nodes form an independent set, but there exist independent sets with more nodes. The blue nodes form a maximum independent set, because all nodes are independent and there are no independent sets of size >5 in the graph.

            Let’s look at an example as applied to banking. The mathematical problem of optimizing a portfolio of financial assets to minimize risk, called portfolio optimization, can be translated into a maximum clique problem [see definitions], which is equivalent to the MIS of the complement graph [see definitions], i.e., the graph where the nodes are connected if, and only if, they are not connected in the original graph. This complement graph is going to be very dense when starting from a sparse graph. In general, that means the complement graph is difficult to encode on a neutral atom quantum computer due to the physical constraints of a neutral atom quantum computer. The cost of mapping and scaling in comparison to classical systems, in addition to the overall integration of classical and quantum processes, are critical factors to assess. 

            Within Capgemini’s Quantum Lab, we have begun two research projects to assess the suitability of neutral atom quantum computers for different applications. The first looks at how to map a problem so it can be run on a neutral atom quantum computer. As these computers can only solve a very specific class of problems – an MIS on a UDG – the challenge is to translate a business problem into an MIS on UDG in order to understand how commercial value can be derived. Several methods exist for this. One of which is a method published by QuEra that translates any MIS problem on a graph to an MIS problem that can be run on a neutral atom quantum computer.1 The challenge with this approach is that it can lead to high overhead and only works if the quantum computer is able to find exact solutions. 

            But do users always need exact solutions? Most of the time, the answer is no. A parcel delivery service does not need to know the exact best route. They just need a good route before the day starts. Even if there exists a route that improves the solution by 0.001%, that might not be worth spending hours or even days of compute power. 

            That is why, in the second research project, we are considering the scaling of the quantum solutions compared to classical approximate optimization solvers. By using QUARK, an open-source benchmarking framework created by software engineers at BMW and an ever-growing community,2 we will create random (unit-disk) graphs and let both the approximate classical algorithm and the (simulated) neutral atom quantum computer attempt to find large independent sets. Then, we measure the (estimated) runtime and quality of the solutions and extrapolate. The goal is to find structures of graphs where the classical methods suffer, but the quantum computer is still able to provide large independent sets.  

            Finally, we will combine the two research projects and define a methodology to assess the potential of neutral atom quantum computers. We will focus on specific business problems, and we will consider the whole (end-to-end) algorithm. This includes translating it into a suitable format for the quantum computer, running the quantum algorithm, and interpreting and validating the results. We will compare the performance of a hybrid (quantum-classical) pipeline with a purely classical one. Factors such as accuracy, speed, scalability, and computational (classical) cost will all be considered. This will help us identify the domains and scenarios where neutral atom quantum computers can offer a significant advantage over classical methods. It will also reveal the challenges and limitations that need to be overcome to realize their full potential. 

            Where to look for good applications 

            So, where do we start to find these applications to assess? Assessment generally begins with either a technology or industry approach. Typically, quantum computing companies start with the technology because they focus first on what they know about quantum computers. Let’s take a deeper dive into this and look at three problem formulations: maximum independent set, graph coloring [see definitions], and minimum vertex cover. 

            Using a technology approach to find applications 

            Maximum independent set 

            It is natural to start with the maximum independent set problem on the unit disk graph that’s been mentioned previously. The goal of the MIS problem is to find the largest set of nodes in a graph that have no connection to each other. An application of this mathematical formulation, for example, is finding the best placement for antennae so that their signals do not interfere with each other. The challenge, though, is to reframe other problems in a similar way. For example, one could try to formulate a traffic optimization issue as an MIS by creating an abstract graph where traffic routes are nodes that share an edge where the routes share a road. The challenge is that this very specific traffic problem is mapped to MIS-UDG, potentially creating high overhead results in the need for more and better qubits to find a solution. These extra qubits don’t contribute to solving a ‘larger’ problem; they are only necessary to map the problem. Furthermore, solutions to the mapped problem might not be actual solutions to the actual problem. This is especially true for the approximate solutions to the mapped problem.  So our challenge is to find formulations of business problems that fit MIS as close as possible. 

            Figure 2 an example of a scheduling problem that can be translated to a graph colouring problem.

            Sometimes, the problem does not require finding one set of nodes but many sets of nodes that are all independent of each other. An example of this would be planning in a production plant, where different tasks can happen at the same time, but not two tasks that require the same resource (e.g., a specialized machine). In this case, the solution to the planning problem is a collection of sets of nodes that don’t share edges. This is an example of graph coloring, wherein a specific color is assigned to all nodes, such that no two connected nodes have the same color as depicted in the figure above. 

            To find a coloring with a neutral atom quantum computer, the problem once again has to be translated into an MIS problem. There are multiple ways to then solve the problem, but they mostly rely on the fact one color is an independent set and that the optimal coloring contains large independent sets. With a classical optimizer and the results from the quantum computer, a classical computer can find a “good” coloring with high probability.3 To reinforce this approach, Capgemini’s Quantum Lab built a demonstration that solves a planning problem using neutral atom computers: Link to Demo

            Minimum vertex cover 

            In some cases, you do not want all solutions to be independent from each other; rather, you want every node to be connected to at least one other node in a set. A good use case would be immunization, where not everyone needs to be inoculated to stop transmission. Instead, those who are not inoculated need to be surrounded by those who are immune, as seen in the figure from Wurtz et al.4. This is called a minimum vertex cover problem. Luckily, the problem is equivalent to the MIS-problem. The opposite of a maximum independent set is a minimum vertex cover. Thus, a neutral atom quantum computer can assist in finding a solution. A challenge now is that usually, the minimum vertex cover only becomes interesting in large graphs. For example, a healthcare organization developing an immunization strategy for a country would require a minimum vertex cover on a graph of millions of nodes. That is much larger than can be embedded on the neutral atom computers.  

            Figure 3, an example of a minimum vertex cover problem in an immunization context.

            The risk in taking a technology-approach, is that the focus in on the specific tool (quantum computing) may lead to proposing cumbersome algorithms, like in the traffic optimization example above. A mathematician might view ‘a constant overhead’ as a small problem, but in practice, this might make the quantum approach unfeasible. Furthermore, focusing on the technology may push to ignore other possible solutions through classical computing. 

            Using an industry approach to find applications 

            Good applications of neutral atom quantum computers can also be discovered by leveraging industry expertise, which is the most common approach for industry leaders. However, focusing solely on bespoke challenges within specific industries and then trying to fit in quantum technology as an afterthought could risk not taking ample account of the technological constraints. This is why ‘quantum Gen AI’, ‘massive parallel computing’, and ‘quantum big data’, make no sense from a technology perspective. 

            A more refined industry approach in the context of neutral atom quantum computing is to look for optimization problems in an industry (e.g., known NP-hard problems) and see if there exists a way to transform them to MIS on UDG. However, this method also has a flaw, albeit more subtle and often overlooked, which is that there could be a structure in the problem that could be exploited by classical computers. This exploitation can result in a good (sometimes approximate) solution in sub-exponential time with a classical computer, which means that the problem is solved. In these cases, trying to tackle a problem with a quantum computer will almost certainly not result in the favorable performance for the quantum computer. While this might result in solving their problem, there might not be a quantum advantage.  

            Capgemini’s Quantum Lab approach 

            Given that the technology approach often overlooks the industry’s needs and that the industry approach often overlooks the technological constraints (of classical and quantum computers), there is a chasm between the two worlds. Capgemini’s Quantum Lab tries to bridge that gap by explicitly looking for classically hard problems where a solution will deliver value. Capgemini’s strong industry expertise and close relationship with industry leaders allows us to test our applications continuously.  

            Closing 

            In conclusion, the journey to finding good applications for neutral atom quantum computers is iterative and collaborative. It requires a deep dive into both industry challenges and quantum computing theories, with a focus on practicality, scalability, and cost-effectiveness. While the use cases are not fully known, ongoing research and collaboration between industry and quantum computing experts are vital for uncovering them. Establishing a robust ecosystem is crucial for the development and identification of practical quantum applications. The partnership between Capgemini and Pasqal exemplifies the commitment to this exploration and the promise it holds for the future of computing. 

            Definitions

            • Maximum independent set problem: The maximum independent set (MIS) problem is a combinatorial optimization problem that asks for the largest subset of nodes in a graph such that no two nodes in the subset are adjacent. This problem has applications in various domains, such as wireless network design, social network analysis, and bioinformatics. However, finding the maximum independent set of a large graph is computationally hard, and classical algorithms may take exponential time to solve it. Neutral atom quantum computers offer a potential speedup for this problem by exploiting the dynamics of the Rydberg blockade.
            • Rydberg blockade: The Rydberg blockade is a quantum phenomenon that arises when atoms are excited to a highly excited state called a Rydberg state. In this state, the atoms have a very large electric dipole moment and experience a strong dipole-dipole interaction. This interaction creates an energy shift that depends on the interatomic distance and the angular momentum of the Rydberg state. As a result, two atoms cannot be excited to the same Rydberg state within a certain distance, which is called the blockade radius. Thus, the excitation of one atom to a Rydberg state blocks the excitation of nearby atoms. This effect can be exploited to create entanglement and implement quantum gates between atoms. 
            • Unit disk graph: A unit disk graph (UDG) is a type of geometric graph where the nodes represent points in the plane, and two nodes are connected by an edge if and only if the distance between them is at most one. UDGs can be used to model wireless networks, where each node has a fixed transmission range of one unit. Neutral atom quantum computers can find MISs on a UDG by using the Rydberg blockade, because two atoms cannot be in the Rydberg state at the same time if they are close to each other. 
            • Maximum clique problem: The problem of finding the largest subset of nodes in a graph that are all connected to each other. 
            • Complement graph: A complement graph of a given graph is a graph that has the same set of nodes but the opposite set of edges. That is, two nodes are connected by an edge in the complement graph if and only if they are not connected by an edge in the original graph. 
            • Graph coloring problem: A graph coloring problem is a combinatorial optimization problem that asks for the minimum number of colors needed to assign a color to each node of a graph such that no two adjacent nodes have the same color. Graph coloring problems have many applications in scheduling, map coloring, register allocation, and Sudoku puzzles. Finding the optimal coloring of a graph is NP-hard, which means that there is no efficient algorithm that can solve it in polynomial time for any graph. 

            Nguyen, M.-T. et al. Quantum Optimization with Arbitrary Connectivity Using Rydberg Atom Arrays. PRX Quantum 4, 010316 (2023).

            Finžgar, J. R., Ross, P., Hölscher, L., Klepsch, J. & Luckow, A. QUARK: A Framework for Quantum Computing Application Benchmarking. 2022 {IEEE} International Conference on Quantum Computing and Engineering ({QCE}) (2022) doi:10.1109/qce53715.2022.00042.

            da Silva Coelho, W., Henriet, L. & Henry, L.-P. Quantum pricing-based column-generation framework for hard combinatorial problems. Phys. Rev. A 107, 032426 (2023).

            Wurtz, J., Lopes, P. L. S., Gemelke, N., Keesling, A. & Wang, S. Industry applications of neutral-atom quantum computing solving independent set problems. Preprint at http://arxiv.org/abs/2205.08500 (2022).

            Meet the author

            Camille de Valk

            Quantum optimisation expert
            As a physicist leading research at Capgemini’s Quantum Lab, Camille specializes in applying physics to real-world problems, particularly in the realm of quantum computing. His work focuses on finding applications in optimization with neutral atoms quantum computers, aiming to accelerate the use of near-term quantum computers. Camille’s background in econophysics research at a Dutch bank has taught him the value of applying physics in various contexts. He uses metaphors and interactive demonstrations to help non-physicists understand complex scientific concepts. Camille’s ultimate goal is to make quantum computing accessible to the general public.

              Resilient and sustainable supply chains for A&D
              How to solve challenges and seize opportunities

              Gilles Bacquet
              17th July 2024
              capgemini-engineering

              There is a growing demand for greener travel, and more opportunities to serve increasingly security conscious countries across the world. Investment in the right technical tools, and a coordinated effort can help us to build a better supply chain for the A&D sector.

              The Challenges

              Aerospace and defense are obviously interconnected. That’s why we often group them together as ‘A&D’.

              Both sectors today are experiencing separate growth in demand for their products. The COVID-19 pandemic’s after effects are still being felt in both industry’s supply chains. Shortages of essential materials, like titanium, cause production delays, as do labor shortfalls in some areas. Constraints in the supply of specialist electronic components, like chips, is another mutual problem.

              Like much of the world, both industries are weathering overall inflation in the costs of raw materials, transport and labor. And, both industries feel increasing societal and shareholder pressure to adopt more sustainable practices in production and supply chain operations, though to a lesser extent in defense.

              A&D’s OEMs and tier 1 companies often share the same suppliers and cross industries – which can lead to the intractable problem in which OEMs are able to source parts, but tier 1s can’t, resulting in these tier 1s being unable to service OEMs. This pressurizing of suppliers can create a kind of ‘cannibalization’ within the industry for certain scarce components and materials – creating bottlenecks and delays that harm the industry as a whole.

              Aerospace and defense also have their own unique challenges and opportunities, some of which are outlined below.

              Aerospace

              In aerospace, up to 80% of a product’s value can come from suppliers. The sector has huge levels of complexity and safety critically in its products. For example, a commercial airliner may comprise millions of individual parts – and the lack of just one of these parts is enough to ground that plane. But the lead time for parts is unusually long, two years in advance in some cases – making forward planning a challenge.

              Aerospace’s current problems range from book orders full for years, manufacturing rates that never exceed rate 75 (75 units a month) for single aisle aircraft, major quality issues – covered to a damaging extent in mass media – and a weak supplier ecosystem (eg. one business analysis forecasts one A&D supplier bankruptcy per week in the next few months).

              But despite all this, the industry needs more airframes – one prediction is for another 32,000 planes in the next 20 years (whether new build or upgraded/retrofitted). Commercial demand appears to be returning to pre-pandemic levels for travel, and on a global scale, demand continues to grow – especially amongst developing countries – even though Europe is seeing a slight drop in short haul flights.

              Defense

              Defense has its own set of challenges and opportunities. It too is seeing a rise in demand – this is driven, unfortunately, by recent conflicts and shifts in geopolitical circumstances and defensive postures. Not only for new systems and platforms, this demand spike also includes repairs to systems currently being fielded and the constant need for spare parts.

              Despite these opportunities, the industry must also endure political uncertainties that affect demand (and to a lesser extent, supply). Defence priorities and spending could change significantly, depending on the results of various elections that are happening at the time of writing. Defense often thrives in periods of instability, but instability must be built into the industry’s forward planning for its supply chains.


              Another challenge faced by defense is to anticipate the future and develop the next generation of equipment that will take us in the 2030s and beyond  (eg. 6th generation fighter aircraft, currently under development in various countries). These new capabilities will require a new supply chain, supporting the introduction of new technologies from suppliers that aren’t yet delivering parts to A&D.

              What to do now

              Despite the challenges, there are several steps that we can take to make the most of A&D’s supply chain situation using supply chain quality management (SCQM). 

              Strengthen our supplier relationships

              Each supplier has separate issues and, as such, each supplier must be managed differently. But, to be blunt, in the current period of scarcity, favorite buyers get favorable treatment. The goal then is to get suppliers to prioritize you when things are scarce.

              Smaller companies don’t necessarily have the knowledge or digital tools to integrate with the supply chain. You can help. This can involve local teams, emphasizing a collaborative approach and personal relationships with your supplier. You can offer supplier training and development programs to upskill their people using the unique experience and knowledge you have developed.

              You can also strengthen your new supplier onboarding, helping new ones to get verified, tackle the appropriate industry standards and overcome the many compliance hurdles of A&D. Contracts can be restructured to incentivize timely deliveries and reward good treatment from suppliers.

              Diversify our suppliers where possible

              The second approach is to diversify, accepting that this can be impossible in many cases due to the current state of the supply chain. Diversifying allows us to support new suppliers and spread risk, considering the volatility of the industry.  The nearshore/friendshore trend is seeing a rearrangement of supply chains , so developing new local suppliers (or ones in allied countries) will help in complying with certain trade policies, particularly in defense.

              Encourage and demonstrate sustainability

              As reported in our blog on sustainability in supply chains, European parlement approved in may 24 a directive for a new supply chain lawknown as ‘The EU Supply Chain Act’ or ‘Corporate Sustainability Due Diligence Directive’ (CS3D). C3SD is intended to address the sustainability problems faced (and caused) by modern supply chains.

              So, as social attitudes and the regulatory landscape change, sustainability is no longer just an ecological concern, it is a business imperative. Inbound and outbound supply chains (scope 3) are well understood to be a major contributor to company emissions – which means steps must be taken now.

              To this end, you should first conduct a detailed ESG evaluation to gain oversight and baseline your current situation and emissions. After this, you can explore ways of embracing the circular economy (where possible), choose greener transport modes within applicable parts of the supply chain and incentivize your suppliers to increase their own sustainability – helping them to do so when you can.

              Improve our supply chain awareness and supplier management

              Knowing the state of our suppliers (and our supplies) is a key part of increasing the resilience of our supply chains. It requires detailed risk assessment, using regression modeling, based on disruptions and accounting for (increasingly frequent) ‘black swan’ events – eg. financial crises pandemics, war – all of which we have endured in recent times.

              This requires dedicated scenario/contingency planning, along with the use of methodologies like the Eight Disciplines (8D) approach. Originally developed at Ford Motor Company – 8D can be used for supply chain problem identification and solving. Combine this with ‘old fashioned’, regularly updated supplier healthchecks/scorecards, conducted by your local team on supplier premises, and, where you find shortfalls, look to help your suppliers improve.

              All of this can be supported and driven by AI/ML enabled big data analytics to predict demand and inventory requirements, along with smarter logistics, eg. real-time tracking via RFID/IoT on all deliveries. The situational awareness granted by such tracking allows us to develop digital twin models of the supply chain – gaining visibility, whilst simulating and improving the supply chain (eg. through route visualization and optimization).

              Conclusion: protecting (and growing) the supplier ecosystem

              A&D’s supply chain is key to its success and onboarding new suppliers into this ecosystem is not easy. In many cases, switching suppliers isn’t an option either, due to the requirements of certification for new suppliers, how few companies are specialized enough to even do this work, and how unstable, as of late, the industry has been.

              Instead, we must work together to create a more resilient and sustainable supplier ecosystem that is capable of withstanding the volatility of our time, and taking advantage of its opportunities. There is a growing demand for greener travel, and more opportunities to serve increasingly security conscious countries across Europe and the rest of the world.

              The supplier ecosystem and the IT landscape are fragmented, but they don’t have to be. We already have the tech and tools to increase visibility, continuity and reliability – we just need to do the significant amount of work required to make it happen. Investment in the right technical tools, and a coordinated effort can help us to build a better supply chain, and a better A&D sector.

              If you are currently facing delivery disruptions, if you need to ramp up your supply chain to meet changing demand or implement sustainable initiatives, we can help. Capgemini has years of experience helping companies across sectors and countries with supply chain quality management, along with access to some of the world’s leading experts in the subject. To find out more, contact our expert.

              Meet our expert

              Gilles Bacquet

              Senior Portfolio & Product Manager, Resilient & Sustainable Supply Chain offers owner
              Gilles is a Production & Supply Chain engineer and has joined Capgemini group in 2001. Starting as consultant expert in Supplier Quality Management for Automobile & Aeronautic, he has extended his responsibilities in creating Supply Chain offer and developed business oversea. He is today leading Resilient & Sustainable Supply Chain offers for Capgemini Engineering.

                Retail media networks – the future of digital advertising

                Abha Singh Senior Director, Capgemini Business Process Outsourcing
                Abha Singh
                Jul 16, 2024

                Building strong relationships with key partners to create a robust retail media network ensures you continue to deliver enhanced value to your customers and brand partners.

                According to McKinsey, retail media networks are currently worth $45 billion and are projected to reach $100 billion in three years – making them the next big thing in advertising.

                This rapid shift is transforming how brands interact with consumers and how retailers generate new revenue streams by monetizing their shopper data. But what are these networks and what’s fueling their growth?

                What is a retail media network?

                A retail media network is an advertising platform managed by a retailer to manage its ad inventory and control the placement of ads from various brands across its owned channels (digital, in-store, etc.) and other paid media channels. This includes ads and videos on its website, app, and in-store digital displays and banners.

                These networks are made up of three main components: the advertiser (the brand or media buyer), the publisher (the retailer owning the digital and physical points of purchase), and a technology platform that connects them.

                Fueling retail media networks

                There are several factors currently driving the growth of retail media networks. These are:

                • The rise of ecommerce – the shift to online shopping has given retailers access to a wealth of first-party data, including consumer purchase behaviors, preferences, and demographics. This enables them to target consumers and create new revenue streams by monetizing their advertising – all while reducing spend
                • Cookie deprecation and the need for first-party data – with Google’s deprecation of third-party cookies, brands are seeking data collection and targeting alternatives. Retailers, armed with extensive first-party customer data, are well-positioned to fill this gap, which encourages them to collaborate closely with retail media networks to create more personalized advertising
                • The disruption of traditional channels – traditional advertising channels, such as TV, have been shrinking for several years, and recent declines in ad revenues for established platforms like Meta and Google signal significant disruption in the rapidly changing advertising ecosystem. However, retail media networks provide brands with the ability to display ads to high-intent consumers closer to the point of purchase, which helps increase conversion rates
                • The presence of lucrative margins – the prospect of setting up a retail media business is highly lucrative for retailers as it can significantly impact their profitability. For example, Amazon’s advertising revenues grew to $46.9 billion in 2023, an increase of over 24% compared to the previous year
                • The rise of non-endemic advertisers – beyond traditional retail brands like CPGs, non-endemic brands are also benefiting from retail media networks. Walmart Connect, Walmart’s ad platform, has started offering offsite media to brands that do not sell at Walmart but offer complementary products and services. This approach positively impacts businesses outside of the retail sector such as automotive, entertainment, financial services, fast-food, and travel providers.

                Collaboration builds successful retail media networks

                In conclusion, while the potential of retail media networks is immense, it’s still very early days. Retailers need to build an entire ecosystem of capabilities from media operating models and streamlined workflows to the right partnerships with agencies, all while providing a personalized experience for shoppers and enhanced campaign performance for advertisers’ retail media spend.

                Even with a robust process and advanced technology, scaling media planning, activating many channels simultaneously, and measuring across multiple advertisers seamlessly remains challenging for many retailers. The increasing deployment of Gen AI and intelligent automation solutions can help accelerate speed to market for campaign activations and provide tangible results for brands and retailers.

                Building a robust and successful retail media network is a journey. Nurturing strong relationships with key partners is an effective way of ensuring you continue to deliver enhanced value to your customers and brand partners.

                To learn how Capgemini’s Connected Marketing Operations is helping Albertson’s Media Collective leverage intelligent process automation and Gen AI to create ad formats, bolster media planning capabilities, and build precise audience segmentation, contact: abha.singh@capgemini.com or isha.b.gupta@capgemini.com

                Meet our experts

                Abha Singh Senior Director, Capgemini Business Process Outsourcing

                Abha Singh

                Senior Director, Capgemini Business Process Outsourcing
                Abha drives large transformation and consultative sales, presales, and marketing projects for Capgemini’s clients, bringing innovation into the core of every area of her work.
                Isha Gupta, EU GTM Lead, Marketing and Communications Services, Capgemini’s Business Services

                Isha Gupta

                EU GTM Lead, Marketing and Communications Services, Capgemini’s Business Services
                Isha Gupta is a subject matter expert and go-to-market leader for marketing services, with over 15 years of experience in the digital marketing ecosystem. She is an expert in the field of marketing transformation, specializing in paid media, martech, and performance marketing.

                  Collaboration, meet acceleration: How to bring together digital threads for faster, more efficient, end-to-end engineering

                  Scott Reid
                  Jul 16, 2024

                  Complexity has become a bit of a buzzword. It seems wherever you look, every business is promising to simplify complexity, without really specifying which bit. After all, what in life and business today isn’t complex? However, complexity is the genuine state of play for the aerospace & defense industry. In battling supply chain issues, cost constraints, rapidly evolving technology, lack of cross-industry and inter-departmental cohesion, and constant changes, ‘complex’ may feel like the only way to describe the many challenges your business is facing.

                  As a Chartered Engineer with a background in defense, it’s these challenges my clients enlist my help with every day. And thankfully, the solutions aren’t too complex. If you’re heading to Farnborough International Airshow (FIA) this year, you can experience live demonstrations of them at Capgemini’s booth or chalet. Plus, my esteemed colleagues and I will be on-hand to show you how bespoke versions of them can be applied to your business.

                  You’ll find more information on our presence at FIA at the bottom of this blog. But before then, let’s break down the answers to the industry’s biggest questions.

                  How can technology help my business balance cost constraints, innovation, and quality, with increasing product complexity?

                  If this challenge resonates, it’s because much of the industry is facing the same thing. Balancing cost constraints with the need for innovation and quality is a perpetual challenge, especially in a highly competitive market. Simultaneously, the products we develop are increasingly complex and we must quickly adapt to them to enable effective and efficient development and management across the full product development lifecycle.

                  The key to solving this? Firstly, Systems Thinking. A way of working that requires probing challenges, asking questions, and interrogating what systems need to achieve before jumping straight to implementation. Taking this “big picture” view of connected wholes rather than isolated parts enables you to break down traditional siloes and is core to efficient product development.

                  Moving forward, by combining Systems Thinking with advancements in digital technologies (both tools and techniques), we can start to address our high-level problem, drive-up efficiency improvements, and manage our products as a holistic system. There are two relevant approaches here, which you can see in action at our FIA booth or chalet.

                  1. Model-Based Approaches. This is essential for advancing traditionally document-centric approaches – which are entrenched with inefficiencies and inaccuracies – and opens the door to more advanced modelling and simulation capabilities. It enables the linkage of models to create a system that understands its boundaries, its contradictions, and can communicate the reasons for its design changes. It is the bridge between requirements and low-level high-fidelity technical models created by domain specialists. Visit us at FIA to: Discover how this approach can help you manage complexity, reduce risk, and increase productivity with a demonstration of C-Pulse – a Medical Transportation Drone, developed from scratch by Capgemini from a Model-Based Systems Engineering (MBSE) model.
                  2. Digital Twins. This term can mean different things, including a digital twin’s simple iteration: attaching several sensors across a system to get a snapshot. However, at Capgemini we always push to utilize MBSE at the core, then expand on it to link up different technologies within the system. Ultimately, it allows us to use the digital twin across a product’s lifecycle and gain a continuous, overarching view, rather than one moment in time. Visit us at FIA to: Understand how digital twins can help you can make data-driven decisions, enhance customer experiences, drive critical efficiency, reduce carbon footprint and uncover new revenue opportunities, with a demonstration of our RT3D airport operations platform – a full digital twin of a North American airport. Learn more about MBSE and Digital Twins here.

                  How do I break down siloes in my organization to improve collaboration?

                  Lack of collaboration is a prevalent issue across the A&D industry; not only between organizations but between in-house engineering domains. Though it’s common to organize teams into capabilities, this approach creates siloes between them – separate data, separate tools, separate simulations and testing – which only drives inherent problems and inefficiencies across their work. Engineering is multi-disciplinary and no one team is more important than the need to produce a seamlessly integrated product to meet our customer needs. 

                  So how do you go about bringing them together? Digital technologies play a big part in enabling and enhancing collaboration; digital continuity breaks down silos by bringing people together not just in ways of working, but also in the tools and data.

                  What can often happen in siloed teams is that each works on their own element of product development with success, but when those elements come together in the physical testing stage, they fail. Our goal, conversely, is to ‘fail fast’ – increasing early simulation and verification activities so we can quickly identify and fix the problems, allowing us to focus spend on the development and management of our products rather than fixing problems later in the lifecycle.

                  Visit us at FIA to: Explore the possibilities of reducing component production time by bringing together specialists from across business functions, through our component design demonstration.

                  What part does agility play in all of this?

                  The thread that runs through these challenges is change. There will always be change – whether that’s in technology, capability, user expectations, customer requirements, or anything else on the extensive list of variables. As an industry, we must enable our teams to adopt to and respond to change, instead of spending all their time trying to manage it. Faced with a change in design requirements? Ask yourself how you can accept it, and feed it into the solution, instead of pushing back on it and sticking with what’s always been done.

                  Our industry requires an approach to engineering that welcomes and embraces change, and agility is crucial to adopting that. Combining agile philosophies, a Systems Thinking approach, and collaborative engineering, backed up with Digital Continuity across the lifecycle, we accelerate our engineering capabilities and drive the resilience needed to adapt to change.

                  Capgemini’s vision for FIA, collaboration, meet acceleration, wasn’t devised simply because it rolls off the tongue. We truly believe if you can collaborate, you’ll combat your, and the industry’s, toughest challenges, and together we can accelerate towards an intelligent future that is connected and sustainable.

                  Learn more:

                  Digital Continuity for the Aerospace Industry

                  Digital Twins in Aerospace and Defence

                  Intelligent Supply Chain for the Aerospace and Defence Industry

                  Lifecycle OptimiZation for Aerospace and Defense

                  Meet the author

                  Scott Reid

                  MBSE & Digital Twin Engineering Lead
                  Scott is a Chartered Engineer specialising in Systems Engineering & MBSE, with a background in defence. He now leads a dedicated MBSE team focussed on transforming organisations to adopt MBSE as well as implementing it on major programmes across many industries.

                  Andrew Hawthorn

                  Chief Solution Architect, High Integrity Systems and Intelligent Industry
                  ‘’The railway sector is under enormous pressure to change and is investing heavily in digitisation to improve passenger experience, increase automation and reduce operating costs.’’

                    A new regulation for sustainable and eco-responsible products paves the way for battery passports

                    Emmanuelle Bischoffe-Cluzel
                    Jul 11, 2024

                    Soon (July 18, 2024), the new Ecodesign for Sustainable Products Regulation comes into force across the European Union (EU).

                    This regulation aims to significantly improve the circularity, energy performance, and environmental sustainability of products on the European market. It represents a major step forward in protecting our planet, promoting more sustainable business models, and strengthening the competitiveness and resilience of the EU economy. The new regulation also provides the groundwork for the introduction of battery passports – a major enabler of automotive sustainability.

                    Framework and objectives

                    The new regulation replaces Directive 2009/125/EC and, as the EU says, “establishes a framework for setting ecodesign requirements for various product groups.”

                    The objectives of the regulation include:

                    • Improving the durability, reusability, repairability, and energy efficiency of products.
                    • Reducing the presence of harmful substances in products while increasing their recycled content.
                    • Facilitating the remanufacturing and recycling of products.
                    • Creating rules regarding products’ carbon emissions and environmental footprints.
                    • Increasing the availability of information about product sustainability.

                    A key innovation in this regulation is the introduction of a digital product passport (DPP). This will provide information on technical performance, materials, repair and recycling capabilities, and environmental impacts throughout the product lifecycle.

                    The information from the DPP will be accessible electronically, enabling consumers, manufacturers, and authorities to make decisions that take into account product sustainability, circularity, and regulatory compliance.

                    Events

                    Mondial de l’Auto 2024

                    Meet Capgemini at one of the largest European Automotive Events, celebrating its 90th edition and 126th anniversary.

                    Register now

                    Digital battery passport

                    Batteries will be the first product group where DPPs will be mandated (from 2027). As we all know, batteries play a central role in sustainable transport and energy transition. They power electric cars, trucks, and other forms of transport, as well as storing the intermittent energy supplied by renewable sources.

                    To decarbonize our world, an innovative approach to batteries is needed. There are two major objectives here:

                    • Securing Europe’s supply of batteries. This requires consideration of the entire value chain, most of which will in future be located within Europe, with domestic mining and the establishment of major recycling projects.
                    • Ensuring that batteries placed on the EU market meet the highest standards in terms of carbon footprint and social and environmental sustainability.

                    The digital battery passport will provide transparency on production conditions, usage history, and crucial information for the repair, reuse, and recycling of batteries. The passport will integrate and disseminate information from across the partner ecosystem, and from all the way along the value chain, from mining to end of life. In this way, it will help to guarantee that batteries meet the highest standards of sustainability throughout their lifecycle.

                    Implementing the digital passport concept for batteries will challenge automotive companies, but the challenges can be overcome by collaborating with the right partner ecosystem and leveraging the latest data-driven tools. The effort will be worthwhile because, compliance apart, battery passports represent a big step forward for automotive sustainability.

                    Sustainability in automotive: From ambition to action
                    Capgemini Research Institute

                    Sustainability in automotive: from ambition to action

                    capgemini-research-institute
                    Capgemini Research Institute

                    Automotive supply chain: Pursuing long-term resilience
                    Podcast

                    Driving the future – a podcast series

                    Author

                    Emmanuelle Bischoffe-Cluzel

                    VP – Sustainability Lead, Global Automotive Industry, Capgemini
                    Emmanuelle Bischoffe-Cluzel offers practical IT and engineering solutions to support automotive sustainability. She has 30 years’ automotive industry experience, gained with a global automaker and a tier 1 supplier, in roles ranging from manufacturing engineering to business development. She holds four patents relating to engine assembly.