RPA and Lean, A Must

I have read a number of papers and articles on the reasons RPA projects struggle or fail and have personally witnessed a number of struggling RPA initiatives.  For anyone that has seen a mature RPA tool in action, you will be surprised to hear that failure numbers as high as 50% are being reported.  RPA tools are very easy to use, lightweight on the network, and easy to adjust as needed.  I have seen very capable RPA bots built in less than a day and complex bots built in a couple of weeks.  So why are so many RPA initiatives struggling?

One of the key reasons for RPA project struggling and even failure is a lack of true process expertise.  Process expertise is needed in the assessment, design, and implementation of bots.  It is also needed for the bot building process itself.  To solve this problem, enlightened organizations are adopting the proven methods and principles of Lean.

In this article, I provide a quick overview of the Lean principles that apply to RPA and how, along with a glimpse into the powerful body of knowledge MSI has developed in our Lean Automation practice.

Lean Facilitation Skills: A true Lean Master has conducted dozens and sometimes more than one hundred Lean process events of various types from Lean Strategy (a.k.a. Hoshin Kanri) to Lean Design through basic Lean Improvement events.  These Lean facilitation skills are vitally important in the RPA bot building process to move swiftly to the ideal process and to get stakeholder agreement on what a bot is supposed to do and how.  In our experience, stakeholder agreement on the steps a bot will take is often the most time-consuming task in the bot life-cycle.

Value Stream Analysis: The ability to define and assess value is a vital first step in the creation of a bot project portfolio. Understanding value, defining value streams, and the subsequent analysis allows us to create an orchestrated bot portfolio that actually reduces the time from input to outcome.  This is a serious problem with most RPA implementations.  They are speeding up micro level subtasks within value streams that merely create backlogs and do nothing to actually reduce the time to value or increase throughput.  Further, the definition of Value Streams provides us a meaningful basis for measuring the ROI of an RPA initiative.

Lean Process Design and Improvement: Matrix Based Design or Axiomatic Design combined with Lean thinking enables you to create profoundly complete and capable requirements for one or more bots in a single pass when your processes need a serious overhaul. For tweaking well designed processes, Lean process improvement will help identify common mistakes such as batch processing and other forms of waste ensuring that processes being automated don’t just speed up bad processes.

Lean Work Cells: Lean Work Cells (a.k.a. Scrum Teams) should be deployed for the bot building life-cycle to ensure work is conducted at its finest practical increment and focus on production is maintained with no hand-offs. The Lean RPA Work Cell may be the most important Lean method you can apply to your RPA program. Clear accountability for bot production and elimination of bot life-cycle hand-offs is of paramount importance.

Kan Ban: Kan Ban should be used by RPA Work cells to control rate of work in a “pull system”, Work In Process (WIP), and to ensure quality specifications for each project.

One Piece Flow: One Piece Flow should be used to ensure in-process inventories/backlogs are not created, that projects are right sized, and that the workload is balanced across work cells.

Poke Yoke: Mistake proofing should be employed in both the bot life-cycle as well as the process automated by each bot. A Lean expert will be familiar with mistake proofing techniques in the digital world making a significantly more robust system.

Be warned, simply doing the same old stuff and calling it Lean will not deliver results. Hanging post-it notes on walls to map processes is not Lean. You must incorporate actual Lean expertise that is only forged through Master’s level eduction, industry experience, and industrial certification to properly adopt and train Lean methods in your organization.

At MSI, we have 18 years of corporate Lean consulting experience and numerous highly respected Lean experts. We have a similar number of years with process automation using various process automation technologies and with the introduction of RPA, MSI has become a pioneer in the adaptation of our Lean Automation and Process Oriented Design techniques into the deployment and management of this breakthrough technology.

Our Lean RPA framework contains all aspects for Lean execution of a Lean Automation program from the Center of Excellence and strategic integration with the business down to hands on Lean RPA Bot development. RPA is coming to your organization, like it or not. Many organizations will stumble and flail about for years attempting to control RPA and turn it into ROI, while those adopting a Lean Thinking approach will benefit early and often.

A subject for another article is Hoshin Kanri, the Lean approach to operationalizing strategy. RPA programs should promote the use of Hoshin Kanri within their organizations and integrate a bot candidate review process for strategic initiatives within the Hoshin Plan. By integrating RPA into strategic initiatives, RPA can propogate throughout an organization strategically, prove its value, increase the probability of success for initiatives, and increase the measurability of initiatives.

RPA, The Lean Six Sigma Game Changer

If you are a Lean Six Sigma Black Belt, Master Black Belt or another type of process engineer / business process reengineering expert, you need to be aware of and get smart on Robotic Process Automation (RPA) technologies.  As a Master Black Belt that has participated in hundreds of process improvement projects in a career spanning more than twenty years, I have seen process improvements attempted in practically every manner possible with varying degrees of executive support, stakeholder commitment, etc.  This is not a post on the importance of leadership commitment or approach.  Plenty of articles and papers are published on those topics.  This is about a technology that will become part of every business computer user’s world within the next five years and it is targeted directly at creating efficiency.

I can tell you that without question, the most successful Lean Six Sigma projects I have been part of all embraced technology as a means for implementing improvements, measuring performance, and driving continuous improvement.  They often leveraged Business Process Management (BPM) or workflow technologies as a program-wide platform.  While very successful, these projects often take longer and require serious dedication from the process improvement team.  They also require an LSS Black Belt capable of Systems Thinking (not all are) and comfortable with technology.


With RPA this is all about to change.  Even in its infancy, RPA is a user-friendly technology that allows users to build digital assistants that automate pretty much any task performed on a computer.  It is like having a multiple application macro builder that watches what you do and then does it for you, over and over again.  For example, one can very easily build an RPA Robot (a.k.a. a bot) that reads all the emails sent to your inbox on a given day, moves all the ones from your retailers into a separate folder, then identifies all the ones with the word return anywhere in the email, then copies all the return forms into a folder, and then copies all the relevant data from each return into an Excel spreadsheet that totals the amounts, and then emails you or sends you a text letting you know that the daily items return list is ready for action with the total number of items in the list and the total dollar amount.  If you want, you could even have it post the returns to your accounting system.  This would all happen within seconds of launching the bot each day rather than the hours it would take to perform by hand.  Other common applications include collecting invoices from vendors and creating consolidated bills for building tenants or collecting personnel data on a single form and then having a bot post it to numerous internal IT systems rather than entering into each system manually.  Bots can take scanned applications, merge them with electronic applications, and then disseminate the information to other systems or create additional documentation and analysis based on the inputs.  Industry is recognizing tremendous savings already in call centers and large-scale financial service centers.  Corporations at the Intelligent Automation Conference in Austin last week including Kraft, Coca Cola, and John Hancock were reporting saving tens of thousands of man-hours and we are aware of Federal Agencies targeting man-hour work reductions on repetitive tasks in the tens of thousands for FY19.  This is just the tip of the iceberg.

I am personally interested in investigating the application of RPA to the massive wave of FOIA requests submitted to the Government each month.  Imagine if a bot simply read each request and then did a scan of each relevant agency system and file server for all related content and then ran a scan for sensitive content and then placed the information in a staging folder for someone to review prior to release.  The Government annually spends millions responding to FOIA requests and rarely meets statutory timelines.   RPA could be a game changer.

If you are like me, you are probably saying to yourself: “This sounds like task automation, not process automation.” If so, good job.  You are right.  RPA is really more about automation of repetitive tasks, not processes with all of their handoffs, business rules, waiting, rework loops, etc.  Other technologies such as BPM and Workflow platforms are very good for that and RPA bots can operate well underneath a workflow technology.  This does not mean that RPA is not a process tool. What it means is that people will be using RPA to automate tasks in silos, or worse, creating the appearance of process improvement and reducing the need for process improvement projects, because they will be reporting a tremendous reduction in man-hours and cost.  If you have read The Goal or you are a process improvement expert that understands Lean, yield, and throughput, you understand how dangerous this could be if real process experts are not involved.

RPA can and should represent a breakthrough for process improvement professionals around the world.  We can leverage RPA to drive significant savings and efficiencies as mentioned above.  We can clearly articulate and execute implementation of process improvements.  We can reduce the LSS project timeline and drastically reduce the need for training.  We can implement a well-constructed architecture of bots that provide real-time process telemetry and drive continuous improvement.  Heck, we can finally start doing SPC and SQC on soft processes!   We can stop process users from blaming the lack of integration among IT systems as the reason they cannot implement efficiencies.  RPA can represent the tangible execution of Improve and Control phases of the DMAIC methodology while getting LSS projects back to a more agile and focused outcome oriented endeavor like they were meant to be.

Alternatively, RPA can be a major blow to our profession by creating the illusion of process improvement when in fact it is really only isolated tasks being improved.  I find that unacceptable.  Process improvement professionals need to take a leadership role in the adoption of RPA within every organization, they must play a key role if not the lead role on all bot implementations, and must ensure RPA deployment creates a more value-centric and measurable enterprise.  There are already numerous articles posted by experts on the extreme importance of process improvement when implementing RPA, see below.  We as process professionals must head the call.

If you are a process improvement professional here is what you need to do.

  • First, start reading on the topic.  You know how to do that, Google it!
  • Second, go to one of the free online training sites provided by vendors such as UiPath and take their free online training. You can get to the point where you can build simple bots on your own with the free training.
  • Third, start asking around about RPA in your organization and do what you can to make sure your CIO and Chief of Process Improvement are collaborating on the topic.  Your CIO is surely already aware of RPA. They must ensure a process improvement expert is on every RPA project.
  • Fourth, ensure your process improvement shop is establishing the standards and practices for process automation from assessment through control of each bot
  • Fifth, begin a campaign to make RPA a standard part of your continuous improvement toolkit such that all Black Belts are trained and able to use it in projects and events.
  • Lastly, via your gate review or some other project review process, ensure RPA is being considered as a tool for improvement on all existing process improvement projects.

As people see the power of RPA when combined with our knowledge of process and our facilitation skills, the entire practice of process improvement will take a giant step forward.

Here are a few links on RPA implementation. You will see they all refer to process selection, design, and change management.

7 key reasons why Robotic Process Automation can fail

Why RPA implementations fail

8 keys to a successful RPA implementation

Three factors for RPA implementation success

10 step RPA Implementation Guide: Pitfalls & Best Practices

What RPA Governance Model is Right for You: Take the Quiz and See

%d bloggers like this: