What knowledge gaps must you overcome to master process automation?

To master process automation, individuals must bridge knowledge gaps in technical proficiency, application understanding, process analysis, data management, security, error handling, optimization, process re-engineering, stakeholder communication, and cultivate a mindset of continuous learning and adaptability.

In business process analysis, balance is crucial. Tools are valuable, but the heart of improvement lies in engaging process participants. User-Centered Insights: Efficient processes rely on daily users who unveil inefficiencies and enhancements. Enhancing User Experience: Efficiency isn't enough; engaging users ensures their needs are met. Collaboration is Pivotal: Successful transformation necessitates collaboration, fostering ownership and effective solutions. Avoiding Assumptions: Assumptions hinder progress; engaging users provides practical insights, grounding strategies in reality. In conclusion, combining methodology with people-centricity drives substantial process analysis improvement, enhancing efficiency and experience.

Two of the most critical "knowledge gaps" are top management's myopia toward long-term results of process automation, and complacency with the status quo. Myopia toward long-term results is caused by the tactical focus on removing process bottlenecks to gain operational productivity rapidly. Of course, this aim is a good thing, but it cannot be the only focus driving a proper Process Automation strategy because it is a Strategy that companies need for Process Automation. Complacency with the status quo is a very common situation and rarely challenged. Studies and research show that up to 60% of employees' tasks are "work about work", that is tasks required to overcome the limitations of IT systems (data entry; reporting; info retrieval).

A key step before any automation work is started is to get a clear grasp on the exact manual steps that are usually performed, including all logic, decision points, edge cases, dependencies etc. Often the process sequences exist only in the brain of the longest tenure employees, and may not be documented. Manually do the task yourself, document & diagram as you go. Then do it again, & refine. Repeat often enough that you have clarity on what constitutes the core repeatable process which covers 90% of key use cases, and which bits are the exceptions. The core process should be automated to bring value to most of your customers, without cluttering the logic of your solution with edge cases, which would make it fragile and hard to evolve.

?? How do you master process automation? Before you dive into tools & software, you need to understand the process in depth & to a level of detail no one else is familiar with. There is a Japanese business term called “Gemba” which refers to where the action takes place. You need to go to the "Gemba"! Whether that is physically going to a place & observing a process (best) or remotely observing through screen sharing & videos (second best). Once you truly understand what is happening after observing your process (multiple times with multiple people), then you need to create a process map & within it include your metrics and following best practices for mapping. Be mindful to map more than the “happy path". ?? Like to support this

Balanced Tool Selection: Considerations for choosing the right tool: Trends vs. Needs: Prioritize utility over trends. Personal Preferences: Focus on use cases, not personal likes. Incentives vs. Experience: Seek a unified user experience. Must-Have Criterion: Choose a tool that fits your needs. Balance ensures effective tool selection.

There is no company in the world that went out of business due to NOT using automation tools. There is a graveyard of companies that used all the best tools but did not deliver results according the user expectations. Thus, if you defined the end result (or the output) in a detailed manner, the choice of tools should be aligned to that deliverable. Many competitors of yours buy fancy tools and "force" their customers to use those tools because it is easier for them or they got a deal for those BPMs, RPAs, and whatnot. Choose the tools that help you make your customer experience better. Once you start thinking from that perspective, there is not much choice to pick from and it'll be quite easy to do.

There are several families of tools: with code, low code and no code. From my experience many low and no code applications like RPA extension do perfectly the job in some parts. But coding offers a great advantage to automate a more complex process. Many AI solutions existing on the market can provide support with coding, they are a good copilot to proceed with the script development. I definitely recommend writing down specifications before anything else to see if a software functionalities can tackle the request.

The use of tools is a very critical and very important issue. When we have more detailed information about the tools used, and especially the more knows about the capabilities of the tools, the faster and more efficient the processes are carried out.

Choose a Language that has broad support within your org, a large community knowledgebase to consult when things go wrong, and a good supply of candidates in the general workforce. Other factors like run time performance are probably less critical for automation tasks, especially if your business process is asynchronous or long running [cart>pay>ship]. Start with the simplest, and modular approach that will deliver the results, and iterate from there.

I strongly believe that unless you are growing past 200 employees, there is nothing that cannot be done with no-code tools. The moment you realize that you need to know what is Python, etc., you'll become a tech company.

Copilot or blackbox code are two examples of good support with programming and scripting. GPT4 can also provide useful support with programming especially in Java and Python. To define the perfect prompt I also use some prompt generator.

Data is undoubtedly essential for process automation, but not all data is equally useful or relevant. To master process automation, you need to start with a wide range of available data and seek for ways to collect even more data from different (open) sources and formats. This will help you to have a comprehensive and diverse data set that can capture the complexity and variability of your processes and have room for maneuvers during the data evaluation phase if needed. Afterward, you need to clean, transform, and prepare your data for analysis, such as removing outliers, missing values, duplicates, and errors. This standard set of steps benefits greatly if you have the proper data pool to evaluate.

Understand data formats, data transformation techniques, and how to integrate different systems and applications. Mastering process automation is an ongoing journey, and staying curious and adaptable to new technologies and methodologies is crucial. Hands-on experience through projects and real-world applications will greatly enhance your expertise in this field.

If you don't know where to start, start with simple things. Every email's attachment should be saved into the folder. This cloud folder should have a name. Think of your company's data as a large folder cabinet where the documents are stored. You've seen plenty of those in the old movies. Because, one way or another, when you really need to build a serious data pipeline, you'll achieve significant cost savings if the source data is correctly organized.

For some time, large pharma and fine chemical companies have been using AI to discover faster pathways to new molecules and chemical processes. This helps establish viable green chemistry pathways and reduces drug development costs. AI is also used in some chemical companies to achieve faster process optimization of multi-product plant. At Autichem we are interested in using AI to develop faster scale up routes for continuous chemical processes which deploy dynamically mixed flow chemistry equipment. This involves machine learning data capture and developing predictive algorithms to establish faster viable scale up routes. We are working with some of the UK’s most prominent universities in this area.

AI has nothing to do with process automation. AI is a set of technologies that is becoming mainstream extremely fast. Therefore, if you are not a large enterprise, then the thing that you need to know about is this: 80% of the time is spent on data, and 20% is spent on the rest in any AI project. The success of using AI in your company is directly correlated to how diligent you are with your data.

They are very important technological developments. Using these latest developments as much as possible and integrating them into processes produces very valuable results.

Automating a process would often also involve changing the process too. Some standards must be established and there cannot be things left to discretion, if the tasks are to be performed by machines. This often makes some people insecure, as they are already very used to the existing way of working. In general, people would usually not be receptive to a new way of working at first. Such resistance to change is understandable and is important in identifying potential loopholes or important details which has been overlooked. Communication is very crucial to reassure the stakeholders, especially the end users. If you want the end users to be open to learn, un-learn and re-learn, make sure they don't need to do, un-do and re-do.

This is most likely the most important aspect - the empowering of stakeholders and to have a phased outcome. The word is cascaded.

The human phenomenon should be at the focus of all kinds of development and change. Even though perfect systems are designed, the project is unlikely to be successful unless the people who are part of the operation of the system are taken into account or if proper communication is not established.

A critical knowledge gap is "Process Automation Mindset". This mindset focuses on identifying opportunities for process automation, recognizing the potential for efficiency gains, and continuously seeking improvements. Why is it Important? Opportunity Recognition This mindset enables individuals to continuously spot areas where automation can add value. Continuous Improvement Automation is a journey that never ends. Automation solutions must change as needs and technologies change. Resource Allocation This mindset helps evaluate whether a task should be automated, based on factors like complexity, frequency, and cost. Problem-Solving This mindset approach views problems as opportunities to streamline and optimize operations.

Since process automation instills fear among workers as they may turn useless, this is the most important aspect to talk about human side of the change. Automation improves productivity and also creates job but in a different skillset. So, training people is an important aspect. People react to changes differently based on their mindset and experiences. While communicating changes, we should plan our opening remarks in a very careful manner because people have tendency to jump to conclusion based on first few sentences. Rely on data driven decision making and avoid criticizing the old practices. It was indeed right at some point of time. We should take all stakeholder feedbacks and communicate clearly that change is open to few adjustments.

I would like to express my additional opinions on this subject as follows: a. Curiosity, motivation and willing to growth: When you have these kinds of driving forces, you can easily find ways/solutions for systems. b. Security awareness: Safety is one of the top priority issue that we have to consider. c. Project Management: All steps of the process/project have to be managed in terms of Project management diciplines. d. Simply is the best: All operation/process should be constructured as much as simple. When modifications or service activities are needed, it is easy and fast to interven.

In some point if your RPA career you will need to deal with JavaScript and DevOps. There are some actions in web automations which Selenium will not handle. You will need to write some JS code. DevOps you also will need to orchestrate RPA's and provision VM's, on premisses or in cloud.

One gap I see is not able to visualize how the change due to automation will effect individuals both operationally and mentally. Playing out the post automation scenarios with people who are part of the process could help in knowing about the gaps.

It may sound ironical. The biggest knowledge gap is in understanding the current process, which is evolved over a long period of time, with contributions from many stake holders under different circumstances, biases, the then technology limitations, availability of skill sets in the society contributing to that. So, before going for process automation it is always better to run through the current process and revalidate them in the current context. You may find yourself surprised to learn how much it simplified your automation job.

Becoming a process automation master is like embarking on a journey through a technological labyrinth. We must navigate various knowledge gaps that could obstruct our path to efficiency. These gaps encompass a deep understanding of automation tools, proficiency in emerging technologies, and a grasp of intricate business processes. It's akin to learning a complex language, where each line of code contributes to the grand narrative of automation. We embrace continuous learning to bridge these gaps, transforming them into stepping stones toward automation mastery.