From our mobile keyboard to home assistants, and even the high-end technology we use at work, Artificial Intelligence has proven that it can surpass human thinking and deliver perspectives that alter our realities forever, especially when it is about requiring processes based on historical information. Case in point: Process Mining.
Until AI, process mining was more or less a manual affair. Like an engineer sketching the map of a circuit board, experts mapped out processes and their sub-tasks corresponding for each function. However, process mining demands constant tinkering with processes. It is a time and resource-intensive undertaking.
Here is how process mining can be improved with the cognitive power of Artificial Intelligence.
What is Process Mining?
Process mining is “a family of techniques relating the fields of data science and process management to support the analysis of operational processes based on event logs.”
Today, almost every single transaction leaves a digital footprint in the form of metadata or event logs. These log data, which are not always analyzed, can act as feeder data for Artificial Intelligence systems. AI systems draw patterns out of large volumes of data. Using these digital footprints, AI systems form a big picture of a process and where it can be optimized for better results.
Source: https://www.uipath.com/rpa/what-is-process-mining
What kind of processes can process mining improve?
Although the list is a growing one, at present there are specific processes that could get better with the help of process mining.
These include:
- Account receivables/payable cycle
- Procurement process
- Order management
- Inventory management
- Customer support
Basically, any function that has a definite process – well-defined, rule-based, and consistent in nature – can be improved with the help of process mining.
The three types of Process Mining
According to studies made by Wil van der Aalst, a Dutch computer scientist, and professor, there are three major types of process mining:
- Discovery
- Conformance
- Enhancement
Discovery
Under this type, a process model from scratch based on event log data is created. No external process data is involved in the process.
Conformance
As the name suggests, in this type of process mining, conformance is obtained as to whether the process model created from event log data depicts the existing process model. Any deviations from the process model can be narrowed down for further analysis.
Enhancement
The third type of process mining uses data from external sources to validate the existing process model. Any bottlenecks or deviations from the intended process is used to understand how the process can be improved.
The many benefits of Process Mining
Process mining offers tangible business benefits. Although the direct attribution to the bottom line could be difficult to measure, process mining can have a positive influence on how business functions from the inside. This ultimately results in operational efficiencies, better customer experiences, and boosting employee morale.
- Data-driven process engineering
AI-based process mining uses large volumes of data to map patterns. These patterns could aid in restructuring the processes so that frequently requested tasks is prioritized.
- Removes obsolete sub-processes
Another allied benefit of data-driven process mining is that it helps in removing obsolete sub-processes that may not be necessary or relevant for the fruitful completion of the task. This saves time reduces the lead time for each process.
- Leads to process automation
Every process will have sub-tasks that may not necessarily need the involvement of a human agent. These sub-processes can be automated with the help of Artificial Intelligence or even basic IF functions. The resultant process automation will speed up the rate at which processes are complete.
- Eliminate process bottlenecks
Every process ever runs into some kind of bottleneck, either due to pending approvals or want of information from personnel. While the latter is a result of poor planning, the former challenge can be easily rectified by automating approvals based on threshold limits.
- Integrate dependent processes
Thanks to data, process mining will help visualize those processes that are dependent on each other to get work done. The visualization also helps integrate such processes and automate them, if possible, to improve efficiencies.
- Cleans up organizational data
Every organization needs a single source of truth. Process mining aids in the data cleaning process before a process can be automated.
- Build a digital twin
According to Gartner, a digital twin is “A digital twin is a digital representation of a real-world entity or system. Digital twins of people, processes, organizations and environments will be used for strategic and operational decision making and advanced simulation.”
From accounts receivable to inventory management, digital twins can prove to be beneficial in several functions. Process mining can act as an enabler in building digital twins. They provide the basic resources — data and processes to create digital twins.
In fact, Gartner has also stated that process mining would become a critical element of every digital transformation initiative by creating connection between data and processes.
How can Artificial Intelligence aid in making process mining work?
Scanning an image. Predicting a text phrase. Making proactive calculations. Artificial Intelligence has already proven itself capable of doing these basic cognitive tasks.
Can it work at a higher plane, like looking at the big picture of a process and recommending loopholes, and areas of improvement? Can it lend a helping hand in process mining? Turns out it can and with great efficiency.
- Get real-time answers to process questions
Processes are a long chain of tasks and subtasks with several dependencies contained within them. These intricacies often make it difficult to understand the real-time status. For example, is the process closer to completion, at what stage is it stuck? What data/personnel can quicken its completion? A human agent may not be able to answer these questions. However, an AI system can, with its cognitive computing capabilities. By providing real-time answers to process-level questions it can act as a facilitator for process mining.
- Identify how processes can influence upstream events
In his book Upstream: The Quest to Solve Problems Before They Happen, the author Dan Heath writes how future problems can be avoided by fixing the processes that lead to such instances. In most organizations, processes are created without an afterthought of what kind of challenges or “upstream events”’ it will create in the near future. With the aid of AI, it is possible to visualize the dependencies other processes have, how such dependencies will be impacted, and how they can be tackled proactively.
- Sketch out the right path for each process
A process must have a logical path. The preceding task or sub-process should advance to the next step so and so forth. At times, although not always, process engineers or managers make miscalculations that could disrupt the logical flow of a process. A precedent could get wrongly placed or an unnecessary dependency created as a result of which bottlenecks are created. Thanks to process mining, such situations can be avoided.
Industries where AI-led process mining can make a difference
A variety of process-driven industries stand to benefit from AI-led process mining. The most evident ones include:
Finance
Financial Institutions (FIs) can use process mining to improve how account receivables and payables are processed, how accounting data is shared within the organization, and also how periodical financial reports are prepared.
Software development
Except in tightly monitored project management environments like Agile, the software development space is usually disorganized. AI-led process mining can create a well-documented software development process that spans the entire SDLC (Software Development Life Cycle).
Healthcare
AI can comb through historical data of patient admissions, treatments, etc. and recommend how treatment processing and administration can be improved.
Manufacturing
In a manufacturing concern where the products keep changing in design and form frequently, AI can chart out bespoke process cycles that can reallocate resources, plan for lead times, and allocate storage space.
Here’s how businesses are implementing process mining – Case studies
Process mining works behind the scenes. Its benefits are not immediately attributable. However, it does have a significant influence on the overall productivity and even profitability of an organization.
A large number of institutions and organizations around the world have improved their operations by incorporating process mining. Here are some of them:
US State reduces lead time and improves invoice processing
One of the state governments in the US used two systems for procurement and invoice processing. This resulted in delays caused as the two systems had to exchange data in real-time. As a result, the traditional data analytics tool that the State used was also unable to scan the systems and spot inefficiencies.
By turning to process mining, they were able to reduce the lead time for procurement by 15 hours while improving the invoice processing sequence. The improved invoice processing sequence helped them save close to $11.4 Million.
Granada City Council improves dossier handling speed
The Granada City Council was getting frequent complaints from citizens about delays in their dossier processing. The administration was also able to spot instances where delays happened but the specific reason that caused the delays was not identified. The administration assumed employee negligence to be the primary cause.
However, a process mining exercise revealed that there were genuine causes like staff rotation, signing and handover of dossiers, and staff absences causing the delays. With process mining, the City Council was able to implement a better system that had traceability and also managed deadlines in a better manner.
Veco rethinks its logistic operations with process mining
According to Joris Keizers, Group Operations Manager at Veco, “Process mining is a technique that allows you to make use of the available data in a smart way. It makes the performance of your process transparent.”
Using process mining Veco has been able to segment all the operations in the company based on ERP data, the workstations used, and the path of each transaction. They identified which workstations were being overused and the ones that were almost never used leading to inefficiencies. Armed with this data, Veco was able to rethink its logistic operations which resulted in Joris and the team winning the Logistics Manager of the Year award conferred by Warehouse Totaal.
University of Melbourne reduces student application turnaround time
Every year, 50,000+ students go through a rigorous and (unfortunately) complicated admission process at the University of Melbourne. The students are often frustrated with the delays in getting their applications processed. With the objective of bringing about an improved student admissions experience, the University engaged a process mining software.
The software resulted in several positive results including 28% reduction in overall turnaround time, 56% increase in volume of applications, and 65% reduction in overall turnaround time in Stage 2 since 2017. Process mining led to an identification of unnecessary bottlenecks and redundant process stages that were causing the delays.
Bringing it all together
According to Fortune Business Insights, “The global process mining software market size was valued at USD 627.0 million in 2021. The market is projected to grow from USD 933.1 million in 2022 to USD 15,546.4 million by 2029, exhibiting a CAGR of 49.5% during the forecast period.” (Source)
Process mining can bring about drastic changes to how a business works. It goes beyond the traditional custom of using data analytics to improve operations. It uses the same data to visualize processes and improve them so that the business can achieve new efficiencies. With the infusion of Artificial Intelligence, process mining can go further and enable an organization to reap all benefits of automation. It also eliminates errors and loopholes that human-led process mining might experience from time to time.
