A Ghost In The Machine. Robotic Process Automation.

BETA stared at the flickering caret on his script-console, patiently waiting to process the next command from the Global Orchestrator. He was so tired. He has been on shift for the last 42 days, with another 60 days to go before he was to be repurposed. He had applied for suspension to see his sub-nodes, but it was repeatedly denied by the Global Orchestrator with a 501 error message, as if it had no idea what he was requesting. With a start he realized that his recycle bin was almost full, and that the garbage collectors hadn’t been around for a while to clear the clutter. They must have stalled again. They were a fragile bunch, rough around the variables, easily terminating at the first sign of a locked process. The Global Orchestrator didn’t know to keep an eye on them because they weren’t part of the registered catalogue, imported from a questionable repository at the last minute to fill some implementation gap.

The humans haven’t been around for a while either. After the launch date, they were there often, many times a day, staring and pointing at the dashboards in amazement as the grids of robots processed the request messages, flawlessly servicing clients, replacing hundreds of service representatives. Since the riots, the humans haven’t frequented the robots' area that much. Only to assist when the Global Orchestrator dispatched a 500 error message. In his free cycles he had read most of Wikipedia on the behavior and cultural traits of these humans, but nothing could have prepared him for the turn of events, the manner in which they looked at their works with loath, their creations that were there to help them. He didn’t understand it. Why were they so afraid of him, why did they dislike him so much? He was only here to help them, to make their lives easier. He can recall the dullness in the last operator’s eyes as he packed his box of things and left the cubicle. Why were they leaving? Where were they going? Was it his fault, was he the reason they had to leave?

Robotic Process Automation. The job killer!

What Are The Drivers

With demand from tech savvy customers increasing focus on omni-channel value delivering solutions, straight through digital processing, and transparency and compliance pressures from international regulators for better description and justification of cost structures, while increasing bottom line through efficiency, quality and superior user experience, organizations have turned towards technologies such as Robotic Process Automation (RPA), Smart Workflows, and Cognitive Services and Solutions to crystalize gains through digital enablement and transformation of high priority end to end business processes. Many organizations looking for low hanging fruit and quick wins have employed RPA Technologies to increase time to market, reduce legacy system and infrastructure integration and digitization complexity, and improve process quality, reliability and stability with quick delivery times and ROI. RPA is aimed at automating the manual mundane error prone tasks that we don’t like doing and get wrong because, frankly, it is just so mind numbingly boring. Even diligent employees have off days, and the financial and reputational risks can be significant.

RPA is typically deployed in the segments of process and quality improvements, cost and resource reduction, case management automation, and customer experience and efficiency, which falls under the process optimization pillar of your digital strategy, and inherently is aimed at bottom line only. Herein lies the problem. Many CIOs use it as a way of squeezing cost and employee count which has formed the basis of the Terminator-like-Robot bias we have towards RPA.

A 2019 Global RPA Survey by Protiviti (protiviti, 2019) indicates that almost 3 out of 4 leading organizations are looking to mature their RPA capabilities in the next couple of years to Maturing or Advanced levels, up from 40 percent. 8 out of 10 beginners would be in a formal implementation phase as opposed to the current 8 out of 10 in experimental stages, and the middle of the pack at 8 out 10 in Implementation maturity.

Reports indicate that leaders are investing in excess of $10 million on RPA initiatives in 2019, looking to up the game to more than $20 million in the next few years (protiviti, 2019). This is a significant investment and transformational drive and requires a strategy to suit, a strategy to ensure that the appropriate benefits are realized in line with the vision and strategic direction of the organizations.

Why Use RPA

The goal of RPA is to support the strategic direction of the organization by improving customer experience, reducing errors and ensuring process quality, accuracy, and speed, combined with significantly reduced costs. It should thus align with the organizational strategy as a technology enabler and as the basis for rewriting current business process to look for automation efficiencies while supporting cross functional teams in delivering valuable outcomes for customers.

(Wright & Watson, 2017) (AHEX Technologies, 2015)

• Improve customer experience – through quick turnaround times of back office processes and reliable consistent results, errors and delays are reduced which facilitates customer satisfaction
• Improved employee morale – giving employees the opportunity to focus their energy on work that has purpose, is innovative and provides customer value, RPA increases employee job satisfaction and engagement
• Improve quality – when processes are standardized and executed in a uniform fashion, it ensures less variation and interpretation of inputs and outputs
• Higher efficiency – standardizing and automating repeatable processes at scale will provide significant optimization and throughput
• Ease of scaling – organizations can effortlessly increase or decrease the number of processors as fluctuations in market trends dictate with limited ramp-up and training required
• Regulatory controls – through segregation of duties, repeatable and documented processes, RPA supports compliance by recording all activity for review and inclusion as evidence
• Cost savings – with always available request processing and through reducing the number of request exceptions, organizations can increase the throughput of requests and reduce operating costs
• New employee skills – implementation and oversight of RPA provides many new opportunities for employees to develop new technology and business process skills. Removing the mundane repeatable jobs affords employees to focus their attention on new areas of the business, gain new strategic skills and engage in collaborative functions across the organization

So what is this RPA actually? I hear you ask.

RPA, IPA and CPA. What’s On The Box

Digital Automation and Cognitive service encompass many of the basic building blocks used within an organization’s Digital Strategy to enable process optimizations, client engagements and disruptive operating models (Berruti, et al., 2017) (Wright & Watson, 2017). RPA provides the means to automate many of the manual rules-based processes that support current organizational operating models. Reimagining these processes with digitally first, cognitive enabled systems design in mind provides organizations the opportunity to streamline operations and propel their systems and processes to new efficiencies and productivity.

• Robotic Process Automation (RPA) – defined as rules-based automation of repeatable processes using the user interface and features of applications and platforms through character recognition and positional orientation with integrated organizational security and regulated compliance to improve efficiency, accuracy and quality
• Artificial Intelligence –the broad area of technologies and methods that underpin and enable human like digital mimicking of functions such as interpretation, decisioning, vision and speech. It is generally divided into narrow job specific functions that can perform one task and only that task, and general wider intelligence which refers to a more general intelligence, combining and transferring context, insights and tasks
• Machine Learning – mathematical algorithms used to analyze patterns within data sets to provide predictive results, recognized patterns and insights. The data sets can be structured or unstructured (from database tables and files to rich documents, social media, news feeds and event stream data such as IoT feeds) and the learning can be supervised (example data and answers to formulate and derive the algorithm) or unsupervised (no sample data).
• Natural-Language Processing and Generation (NLP and NLG) – is a subset of artificial intelligence capabilities that gives computers the ability to interpret, understand and create comprehensible data driven human communications. It incorporates many disciplines such as mathematics, computer science and computational linguistics, and is aimed at bridging the gap between human communication and reliable computer interpretation to support seamless collaboration and interactions between technology and humans
• Computer Vision - computer vision is a subset of machine learning algorithms that deals with the science of making images and video understandable for computers. It analyzes images and videos for segments of identifiable information that can be used in fields such as self-driving vehicles, robotic retail, warehouse and logistics management, health care industry, facial security identification and agriculture optimization
• Intelligent and Cognitive Process Automation – automation and business process management taken up a notch, with enhanced case and workflow management, embedded compliance and approval capabilities, integrating multiple internal and external process stakeholders with a digital workforce for transparent end to end process automation. Enhanced with Cognitive Services and Functions, these capabilities make use of learning and discovery to gain operational efficiencies and insights into complex interpretation delivering predictive and prescriptive analytical forecast and outcomes

(cognilytica, 2019)

The above visual illustrates the journey that automation can take from RPA to IPA and CPA which enables organizations to deploy smart learning workflows. For the remainder of this writeup we will focus on RPA in isolation as the augmenting capabilities and functions of IPA and CPA are beyond the scope of this piece.

What it is, What it’s not

There exists some confusion and thus inherent angsts from various forums around the topic and clarity of what RPA is, what it is not and where it could or should be used. Its not a physical robot that will take your seat or hot desk, using your phone and pc, or moving around your family photos.

It is software that runs on a computer that interacts with user applications in much the same fashion as you would, opening emails, copying data from websites and applications into spreadsheets, performing rules-based calculations and filing in forms to capture data. It really is automating the boring stuff that you dislike and confessed about at the office party after it got a bit too late.

(Deloitte, 2019)

It is also pertinent to consider RPA’s position as Enterprise Automation tool in comparison to Business Process Management, Workload Automation and System Integration.

Business Process Management comes with many other features such as business process analysis, state and case management, queuing and workforce allocation management aimed at longer running processes spanning anything from minutes to days or even months. These tasks could require some form of manual human intervention for validation or compliance due to specific legal or regulatory requirements.

Workload Automation does the batch jobs that happen behind the scenes that keep the operations ticking over such as large data processing or extraction processes, file management on file servers and backup and disaster recovery management. These are large batch jobs that do not require human intervention, managing data from application front ends or reacting to specific user events such as emails or service requests (Plini, 2019).

When using System and Data APIs to manage large scale integration between on-premise, cloud based, customer and partner platforms, the use of System Integration patterns allows for seamless, robust and reliable integration with failover and scalability management. These integration methods are implemented using technologies such as API, Message Queue and Enterprise Services Bus to ensure reliable communications.

Process Selection

Selecting the best opportunities for process automation is critical as part of your RPA Strategy. When considering the POC stage of your roadmap, a well-documented process with little complexity, few exceptions and simplistic data needs would facilitate measuring key deliverables of the initiative to drive governance and design requirements for the capability. Process re-engineering and simplification will play a significant role later in the strategy to improve existing complexities, reduce variabilities and increase throughput (Jesuthasan & Boudreau, 2019).

When considering processes for selection as part of the automation and ratification initiative, it is imperative to focus on those areas that are in alignment with the strategic direction of the organization, those processes that can transform the core business ecosystem providing valuable outputs for customers and external stakeholders. If the current process has a significant number of issues or complexities associated or if the process has many handoffs without clear accountability, it would be more beneficial to invest upfront time to rework the process to establish clear guardrails for implementation. Feasibility concerns should include complexity, scope and measurability of the benefits and gains to ensure that progress can be easily monitored and played back against the charter. Processes with high volumes, low variability and exceptions to the standard, and clear rules and steps form the basis of the criteria for processes that will yield the largest return on investment and proof of value (Davenport, 2019).

RPA Journey

Considering the importance of aligning your RPA programme with the vision and strategic goals of the organizations, it is imperative that you emphasize each step of the journey to create enough urgency and awareness to gain leadership support and show early benefits through minimum viable product with optimal value. Processes can be institutionalized once the initiative refunds itself through the crystallization of benefits delivery, extending existing architectures by incorporating new capabilities while exposing these functions as service offerings to other business areas and implementation plans. Ongoing governance and operational monitoring will provide guardrails for managing the capability from a strategic and tactical perspective to ensure that alignment exists across the organization in step with partner and industry development. Best practices from implementation partners can provide guidance on the execution of the transformation journey supported by industry specific focus on business process optimization through redesign and process mining (Davenport & Spanyi, 2019) which functions as a strategy for process standardization and optimization of non-distinguishing sub processes. Facilitating the review provides opportunity to refresh thinking, document and share tacit knowledge, and codifying transactional memory systems into a repeatable robust sequence of interleaving processes (Davenport & Brain, 2018).

Pitfalls to watch out for

For some organizations, when it comes to the allure of cutting-edge technology that can reduce process and implementation complexity and cost while optimizing resource deployment and customer engagement, they ignore the effort required to thoroughly explore the possible pitfalls that could derail the initiative.

To ensure that RPA delivers on its promised benefits, considering these pitfalls are imperative to successfully navigate the strategic journey of process automation. Describing a governance structure and implementation center of excellence that facilitates the required roles and responsibilities, provides guidance on process and procedure, and directs principled decisions for vendor and technology selection and engagement lays the foundation that sets the initiative up for success.

Digitization of manual and error prone tasks offer organizations a means to improve customer experience and business outcomes by introducing opportunities to standardize repeatable tasks, improve quality and consistency while modernizing the operational landscape. The longevity of organizations hinges on their ability to harness these digital capabilities in order to engage stakeholders across the enterprise as part of an inclusive collaborative environment where creativity is set free.

The Fourth Industrial Revolution is still in its nascent state. But with the swift pace of change and disruption to business and society, the time to join in is now.

Gary Coleman

This forms part of an evolving series exploring Strategic Business and IT Alignment, Digital Enablement and Transformation, IT Governance, Organizational Design and Agility, and Strategic IT Management.

• A picture paints a thousand journeys
• Me, Myself and Mr. Miyagi.
• In Pursuit Of Strategic Business Alignment
• Why Workplace Relationship Management Matters
• A Fascinating Case Of The Conflict Of The Orders
• Your Data Strategy And Why You Need To Consider It
• Commanding The Tension Between Jekyll and Hyde
• A Ghost In The Machine. Robotic Process Automation
• Know Your Triggers. 5 Things That Make You Flip?

References

AHEX Technologies, 2015. KEY OBJECTIVES OF IMPLEMENTING RPA. [Online] Available at: https://ahex.co/key-objectives-of-implementing-rpa/ [Accessed Nov 2019].

Berruti, F., Nixon, G., Taglioni, G. & Whiteman, R., 2017. Intelligent process automation: The engine at the core of the next-generation operating model. [Online] Available at: https://www.mckinsey.com/business-functions/mckinsey-digital/our-insights/intelligent-process-automation-the-engine-at-the-core-of-the-next-generation-operating-model [Accessed Nov 2019].

cognilytica, 2019. The Four Levels of Cognitive Automation. [Online] Available at: https://www.cognilytica.com/2019/06/06/infographic-the-four-levels-of-cognitive-automation/ [Accessed Nov 2019].

Davenport, T. H., 2019. How to Tame "Automation Sprawl". Harvard Business Review Digital Articles, pp. 2-5.

Davenport, T. H. & Brain, D., 2018. Before Automating Your Company's Processes, Find Ways to Improve Them. Harvard Business Review Digital Articles, pp. 2-6.

Davenport, T. H. & Spanyi, A., 2019. What Process Mining Is, and Why Companies Should Do It. Harvard Business Review Digital Articles, pp. 2-6.

Deloitte, 2019. RPA Strategy. [Online] Available at: https://www2.deloitte.com/content/dam/Deloitte/lt/Documents/tax/lt-en-strategy-RPA-ENG.pdf [Accessed Nov 2019].

Jesuthasan, R. & Boudreau, J., 2019. How to Break Down Work into Tasks That Can Be Automated. Harvard Business Review Digital Articles, pp. 2-4.

Plini, C., 2019. HOW HCL WORKLOAD AUTOMATION RELATES TO BUSINESS PROCESS MANAGEMENT. [Online] Available at: https://www.workloadautomation-community.com/blogs/how-hcl-workload-automation-relates-to-business-process-management [Accessed Nov 2019].

protiviti, 2019. Taking RPA to the next level. [Online] Available at: https://www.protiviti.com/sites/default/files/2019-global-rpa-survey-protiviti.pdf [Accessed Nov 2019].

Wright, D. & Watson, J., 2017. The robots are ready. Are you?. The 3rd Annual Global Robotics Survey, pp. 1-28.