Skip to Content

Robotic process automation an industry perspective

Kaustav Dutta
November 25, 2019

1.     Introduction

To run a business with high efficiency and low cost has always been a challenge. Complex processes, fast changing regulatory requirements, labor-intensive activities, unmanageable volumes, faster time to market – these are some of the key challenges faced by most financial institutions. Operational complexities have reached a state where they can no longer be ignored. To succeed in this complex environment, automation is the key – automation that is more secure, efficient, and scalable. Robotic process automation (RPA) perfectly fits the requirement of process automation because it promises solutions at low cost, minimal disruption, short payback period, and high efficiency.

Businesses across the board are realizing that RPA is the next significant digital transformation that will enable employees to stop working on repetitive tasks. RPA allows employees to concentrate on more value-added initiatives, which are imperative for the bottom line of the firm. This paper aims to provide an industry perspective on RPA and how it is revolutionizing the way companies operate today.

2.   The automation revolution –Creating a digital enterprise

“By 2025, automation technology innovations will assume control over tasks that are now performed by 250 million knowledge workers worldwide, freeing the remaining work force to devote their time and energy to more creative pursuits”  – McKinsey & Company

Throughout history, new technologies have changed the shape of the labor market – both destroying and creating jobs in a process that the economist Joseph Schumpeter described as “creative destruction.” As the maturity of automation, artificial intelligence, machine learning, and cognitive techniques increases, we find ourselves on the cusp of another wave of creative destruction.

According to McKinsey & Company, by 2025, automation technology innovations will assume control over tasks that are now performed by 250 million knowledge workers worldwide, thus freeing the remaining work force to devote their time and energy to more creative pursuits.

The concept of automation can be viewed in a negative light. In the press and analyst community particularly, we often see stories about “predicted job losses” as the “robots take over.” The industry view is more optimistic. Automation is not as simple as replacing man with machine; it is about businesses using advancements in technology to optimize their operations and orchestrate new and innovative ways of working.

2.1   Becoming a digital enterprise

“Automation is a key enabler of the digital enterprise – and your future success”

A primary goal for businesses today is to become a “digital enterprise,” so they can enter new markets, adopt new channels, and compete with new, agile competitors. However, the legacy of old systems, manual processes, and analog thinking can be a barrier. Furthermore, there can be a reluctance to embrace digital change due to a lack of awareness of what the full transformation will entail. Automation is a key enabler of the digital enterprise.

Automation should be viewed positively in terms of what it will enable:

  • Transformation of processes to compete with digital competitors
  • Redirection of resources from repetitive tasks toward higher-value business activity
  • More creative and fulfilling job roles for employees
  • Agility, competitiveness, and new ideas throughout the business.

3.    The spectrum of smart process automation – An incremental journey

“By 2018, 30% of our interactions with technology will be through conversations with smart robots”

                —     Gartner

The spectrum of smart of automation is unnleashing a plethora of opportunities for organizations to on-board their digital journey. With the progressive evolution of automation technology, Gartner predicts that by 2018, 30% of our interactions with technology will be through “conversations” with smart robots.

The figure above illustrates the evolution of process automation and the potential business value it can bring at each level.

The subsequent sections discuss each stage of process automation in detail, with an illustrative example. The example will highlight how incremental progress in process automation has paved the way for a better future and increased addition of business value.

3.1        RPA 1.0 – The era of desktop automation

Desktop automation was the early stages of automation when users automated fragmented tasks using various techniques, such as VBA, Macro, command script, power shell, and message handling with Win32, to ease their workloads.

Illustrative process: account balance inquiry

Key steps involved: 1. Receive email request à 2. Fetch account number à 3. Validate client à 4. Retrieve account balances à 5. Send response in a pdf to client. Input type: Email request from clients (unstructured data) Automation scope: Step 4. Operator uses a macro/shell script that can fetch balance details from source system for the requested account number. The remaining steps are carried out manually.

3.2        RPA 2.0 – The era of robotic process automation

“The global RPA market is expected to reach USD 4.98 billion by 2020, with a CAGR of 60.50%”

RPA is the latest wave of software automation to reach maturity. The global RPA market is expected to reach USD 4.98 billion by 2020, with a compound annual growth rate (CAGR) of 60.50%.

RPA is the automation of business processes through the implementation of set of software technologies that work as a “virtual worker” to manipulate existing application software in the same way that a person completes a process. These “virtual workers” offer improved business efficiency, data security, and effectiveness by mimicking human actions and automating repetitive tasks across multiple business applications without altering existing infrastructure and systems. Enhanced productivity, reduced cycle time, and improved accuracy and compliance are some of the benefits of this technology.

RPA capabilities can be deployed to address a wide range of operational challenges, from improving agents’ handling time to automating of end-to-end processes. Typically, RPA provides two modes of deployment.

3.2.1           RPA 2.1 – Attended robots

  • Attended work involves employees performing tasks (front and back office)
  • The “robot” lives on the agent’s desktop (or their own virtual/Citrix desktop) callouts, rules, automation, copy and paste between systems, etc. are triggered by desktop events.

Illustrative process: account balance inquiry

Key steps involved: 1. Receive email request à 2. Fetch account number à 3. Validate client à 4. Retrieve account balances à 5. Send response in a pdf to client. Input type: Email request from clients (unstructured data) Automation scope: Steps 4 and 5: An attended robot can query the source system and fetch account balances, create a report, and email it back to the client. Steps 1 and 2 still require human judgment and need to be carried out manually by an operator. Further automation can be achieved with few optimizations at the initial steps of the process. For step 2, the requested email template must be converted to  a structured/electronic form. Similarly, for step 3, rule-based client validation needs to be looked at. In this scenario, although all the steps are automated, the bot still needs to be triggered by the operator and if any exceptions are encountered the operator must resolve them on the fly.

3.2.2           RPA 2.2 – Unattended robots

Unattended work requires no human intervention

  • The “robot” lives on a virtual desktop in an “automation hub”
  • Work is triggered by being put onto a work queue which the “robots” subscribe to.
Key steps involved: 1. Receive email request à 2. Fetch account number à 3. Validate client à 4. Retrieve account balances à 5. Send response in a pdf to client. Input type: Email request from clients (unstructured data) Automation scope: Steps 2, 3,4, and 5: As explained in the attended case, with all future optimizations to the process all the steps can be fully automated, and a bot can be built to achieve end-to-end automation using unattended RPA. Unattended bots can be scheduled to start automatically once an email request is received. The unattended bot needs to be deployed in an automation hub and will work 24/7 with no human intervention. The bot’s activities will be monitored and logged in the control center. Any exceptions encountered will be reported on the dashboard automatically and will require human intervention to solve them.

Illustrative process: account balance inquiry

3.3 RPA  3.0 – The era of cognitive and AI automation

Further evolution of RPA incorporates elements of artificial intelligence (AI) and cognitive abilities to mimic human thought processes using machine learning algorithms. It consists of multiple technologies that enable robots to sense, perceive, understand, learn, reason, and infer. Cognitive bots require threshold training to develop intelligent behavior. Greater benefits can be reaped when cognitive bots are combined with the power of cloud computing and cloud APIs.

Illustrative process: account balance inquiry

Key steps involved: 1. Receive email request à 2. Fetch account number à 3. Validate client à 4. Retrieve account balances à 5. Send response in a pdf to client. Input type: Email request from clients (unstructured data) Automation scope: Steps 2, 3, 4, and 5: A cognitive bot can use an intelligent OCR to extract account numbers from emails. It can validate clients based on multiple trained patterns, and it can retrieve account balances and respond back to clients. On top of that, cognitive bots can provide value-added services, such as the ability to handle multiple exception scenarios, or request additional details from clients, etc., – all acquired via training to augment their decision-making abilities.

4.  RPA 2.0 – Taking a closer look

4.1        Can all processes be robotized using RPA 2.0?

Not all processes are fit for RPA automation. The diagrams below provide guidance on how to select processes for RPA automation based on robotics principles.

Processes that show the following characteristics can be considered for RPA:

  • Repetitive manual tasks
  • Rule-based activities
  • High to medium FTE involved
  • Limited human judgment involved in the various activities
  • High to medium volumes of transaction
  • Technically feasible to integrate with supporting systems
  • Fewer dependencies of unstructured data , such as free text emails, scanned or handwritten documents, etc.

“RPA offers a quick breakeven along with improvement in quality, efficiency, and operational agility”

4.2 Why RPA 2.0 – Key benefits

Beyond the hype, RPA offers a quick breakeven along with improvement in quality, efficiency, and control. RPA capabilities are meant to drive operational agility. The key benefits of RPA are highlighted below:

4.3       RPA 2.0 vendors – Key market players

According to Forrester Wave research from February 2017, the RPA vendor offerings are quite diverse. This evaluation of the RPA market should be taken as a starting point only and detailed product evaluations should be based on the organization’s needs, RPA strategy, and use-case areas of interest — for example, attended operations, unattended back office automation, advanced analytics, or Citrix deployments.

“Selection of the right RPA tool should be aligned with the RPA strategy of the enterprise”

Source: The Forrester Wave Report 2019

Based on market insights from leading research companies, Automation Anywhere, Blueprism, and UiPath are consistently positioned as RPA leaders. All these three tools have strong vision, strong automation capabilities, and significantly higher market share. Pegasystems and WorkFusion are strong performers. Workfusion provides strong cognitive features to build intelligent bots, while PEGA Openspan merges robots, analytics, and business process management solution into one platform.

5.           Key Levers To Make Enterprise Automation A Success Story

Successful process automation can be realized in all enterprises by leveraging industry best practices. Such best practices can be broadly classified into four key dimensions, as highlighted below:

 “The success of process automation depends on the ability to leverage industry best practices, customized to meet the enterprise automation aspirations”

Technology

  • Choose the right technology for the right business problem
  • Integrate multiple technology stacks to bring-end to-end automation
  • Break the myth – one platform for all automation needs
  • Choose platforms that integrate well with the existing IT ecosystem
  • Build reusable automation solutions.

Governance

  • Define business participation, steer, and sponsorship
  • Define KPIs to measure business value through automation
  • Centralize the automation center of excellence
  • Unify technology islands
  • Break functional silos.

Change management

  • Drive adoption through workforce transformation and not replacement
  • Drive satisfaction surveys to collate feedback
  • Train users on the new work environment
  • Communicate and collaborate.

Innovation

  • Encourage automation thought leadership and workshops
  • Assimilate and improve from outside-in views
  • Investigate revenue generation avenues from automated solutions
  • Design futuristic processes with a digital mindset
  • Start small with pilots and industrialize.

6.            Conclusion

Today CxOs have understood that to stay competitive businesses must adopt processes that are streamlined, optimal and automated. In short “Automation First” has become a key strategy to meet organization goals. Virtual workforce has become a faster way of delivering business value. Capgemini’s strong professional expertise in RPA is seamlessly helping global customers meet their aspirations.

‘Kaustav Dutta is an Intelligent Process Automation Expert within Capgemini and advises client to understand the automation potential within their respective organization. You can contact him at kaustav.b.dutta@capgemini.com