From Aris to Digital Twin

At the AI Summit 2019 in Munich organized by Handelsblatt and Euroforum, Professor Sabina Jeschke, Board Member for Digitalization and Technology at Deutsche Bahn, surprised everyone by calling for high-performance computers.

Why does the railroad need supercomputers? The answer is very surprising in detail: Sabina Jeschke is planning to build a Digital Twin for rail traffic, i.e. a computer model that simulates the rail network and the trains.

"But I don't want to see a dot on the screen moving from A to B".

she explained at this year's AI Summit in Munich.

It wants and the railroad needs a Digital Twin that simulates the entire train. The simulation of a dynamic system that takes into account and calculates air resistance, track condition, energy consumption, number of passengers, etc. Why?

Public transport is experiencing strong and sustained growth in a short period of time. Experience shows that new rail lines can only be planned and built over decades. To increase capacity for passenger and freight transport, the railroads have only one way out: to move more trains on the existing tracks.

However, as Sabina Jeschke explained, this is not trivial from a technical point of view. Already today, trains on many lines run with minimal safety clearances. If the system is maintained, the purchase of new trains will not help either, because they would simply not find space on the tracks.

As Chief Technology Officer of Deutsche Bahn, she is planning a revolution on the track: artificial intelligence, simulation with supercomputers and a universal Digital Twin are to be used to optimize existing transport routes in a completely new way.

Put simply, Professor Sabina Jeschke does not want to improve what exists, but to redesign it. The rules remain the same. People want to get from A to B as quickly and comfortably as possible. To do this, however, new operating models and organizational and process structures can be developed in accordance with the valid rules.

In her presentation at the Handelsblatt/Euroforum Summit in Munich, Sabina Jeschke used the board game Go to explain what is possible with AI.

Google took over a British AI company and built a Go-playing computer, taught it all the Go literature, tips and tricks. Then they let the computer play against itself for a few months to optimize the Go program.

The result was publicly admired: The Go computer clearly beat the world's best Go player. But this is where the exciting story begins. Google built a second, more powerful computer and gave it only the relatively simple Go game rules.

This computer also played incessantly against itself. The entire knowledge about this complex game was acquired by the computer itself. Over 2000 years of Go literature came into being a second time. Without legacy and defaults, the new Go calculator developed a style of play that has since defeated its predecessor - and the learning continues!

With Aris, Professor August-Wilhelm Scheer created a symbolic language that can be used to describe existing business processes and plan new ones.

"Existing SAP customers are currently asking for a tool to digitize business models, especially as they prepare for the S/4 Hana conversion."

Scheer knows from his company's work.

However, Sabina Jeschke's work raises the question of whether the digitization of existing business models can be sufficient to meet new challenges. Whether a fresh, unprejudiced approach, freed from all old burdens, as shown by Google's second Go Machine, is not the better way?

For Professor Scheer, the version change to S/4 is a good time to unite and realign the strategy with the changing requirements for business processes.

"Aris can play a key role between business and IT in this scenario. With Aris, the processes can be synchronized directly with SAP Solution Manager and provide extensive support in test management."

Scheer explains.

For many existing SAP customers, however, the "Aris wallpaper" on the walls of the conference rooms is too static and one-dimensional, both in terms of a necessary inventory and in terms of planning a digital transformation. The new paradigm here could be "process mining".

One SAP partner that has introduced process mining to the SAP community is the Munich-based company Celonis. The dynamism of the topic and this start-up can be seen in its 2018 financial valuation of over one billion euros. This makes Celonis one of the very few unicorns in Germany.

Which tools are currently used by Scheer Consulting?

"For process modeling and analysis, Aris is currently used by our consultants"

explains Professor Scheer.

In the future, the consulting company Scheer wants to use PAS with its own content Scheer Performance Ready: At Scheer, the Process Automation Suite (PAS) is the answer to the dynamics of digital transformation. In a white paper, Ulrich Storck, Chief Technology Officer at Scheer, explains the path and the function:

Management and automation

Business Process Management (BPM) has undergone a massive transformation in recent years. Starting with classic BPM, which focused on modeling, documenting and mapping ideal processes, it has transformed into a discipline that is increasingly more concerned with information retrieval and process execution (Business Process Automation, or BPA for short). [...]

Today, it is becoming increasingly apparent that this is no longer sufficient: Due to the rapid changes in corporate processes and the diversity and individualization of processes, changes must have a direct and immediate effect on the executed processes, i.e. the process instances.

Such changes to the executed process, i.e., to data or to the flow of instances, are required relatively frequently and provide an immediate benefit, as they affect the currently executed process and thus the value creation. (end of quote)

For lived practice, Professor Scheer explains:

"The combination of methods, tools and content, as we have in Scheer Performance Ready, is an important factor for a successful transformation to SAP S/4 and Hana."

In order to standardize and accelerate corporate processes globally, the Munich-based company Siemens introduced Celonis Process Mining (CPM) back in 2013. The idea behind it: complete transparency in actual processes as well as the possibility to fully exploit optimization potentials.

Siemens users get scalable insights into the flow of processes at any time and can optimize in a timely manner. CPM is a technology that reconstructs and visualizes IT-supported processes live from start to finish.

Business users can analyze and review processes and their numerous variants in real time: Bottlenecks, compliance violations, and inefficiencies are detected and process efficiency is monitored.

Celonis Process Mining runs on one of the world's largest SAP Hana database installations. More than 70 SAP ERP systems are simultaneously connected to Siemens' global Hana environment.

What function can process mining have in the digitization and implementation of SAP S/4? Professor August-Wilhelm Scheer: "As an intelligent and flexible analysis method for business processes, process mining assumes the function of measurement and evaluation in the BPM cycle.

This allows process mining to be used as an assessment and scoping tool in advance of digitization as well as implementation of S/4 to determine scope and priorities and drive harmonization in advance."

Doing the right things, as Professor Scheer emphasizes in the interview. However, after the first implementation steps and throughout the implementation phase, monitoring the processes and improvements is all the more important in order to validate the approach and identify challenges early on: Doing things right!

After going live, process mining supports both the tactical control of processes through continuous measurement in order to detect negative trends at an early stage, and the intervention in individual cases in order to prevent problems from arising in the first place.

"Process mining thus increases both the efficiency and effectiveness of digitization projects in all phases"

knows Scheer as a scientist and practitioner.

Aris versus Process Mining

In 2016, Celonis was awarded the SAP Hana Innovation Award together with Siemens. Celonis Process Mining technology enables optimization of business processes through unprecedented transparency.

It uses SAP Hana to process large Big Data scenarios in real time. Since 2015, SAP has been selling Celonis' process mining technology as part of a global partnership.

Celonis has also been a member of the SAP Start-up Focus program since 2012, which supports start-ups in developing new applications for the Hana platform. Back in 2015, the company received the Hana Innovation Award in the "Most Transformative" category at Sapphire in Orlando.

"We are very pleased to receive this outstanding award for the second time"

says Bastian Nominacher, co-CEO and co-founder of Celonis.

"This shows the enormous potential for large companies in our process mining technology."

Many roads lead to Rome, they say. The starting point for many existing SAP customers is obviously BPM, Aris and the capabilities of SAP SolMan. Future challenges of digitization can probably only be solved with artificial intelligence (Machine/Deep Learning) and Digital Twins.

Along the way, SAP's existing customers can currently make use of techniques such as Robotic Process Automation (RPA) and process mining.

Can Robotic Process Automation be seen as an Aris successor? "Aris and RPA are different solutions for different use cases," says Professor Scheer.

"RPA is the front-end integration for applications, so it's an executing component, Aris is mainly used in business and IT architecture planning. Aris is the planning, RPA is a way of execution."

Robotic Process Automation (RPA) automates individual roles or individual process steps through the use of software robots, i.e., it primarily involves the automation of routine tasks. Aris is a BPM platform that supports digital transformation, among many other topics, i.e. a creative planning and measurement tool.

Robotic Process Automation

Aris can thus be used in RPA projects for strategic preparation, design and measurement. The combination of concepts and tools creates great added value.

Here is a quote from the aforementioned white paper by CTO Ulrich Storck: "Process Automation 4.0 now heralds the next stage of BPA: the complete automation of processes.

Here, human interaction with the process is to be limited to monitoring and handling errors. One of the currently most important topics of Process Automation 4.0 is Robotic Process Automation (RPA).

The central paradigm of RPA is the use of software robots to close existing gaps in process automation. These interact with the applications, analogous to humans, via user interfaces. In addition to rule-based software robots, the use of Artificial Intelligence (AI) is a core aspect of RPA. (end of quote)

"AI stands as an umbrella term for many processes and algorithms that deal with software-based intelligent behavior and analysis and machine learning"

says Professor Scheer.

"AI will thus play an increasingly important role in many areas of companies in the future, including and especially in business process management. Its use will be particularly interesting in the creation and development of new and smart business models as well as the intelligent - i.e. AI-equipped - execution of business processes. This is made possible by the ever-improving algorithms and the almost unlimited computing power of cloud computing."

Organizations today, with millions of jobs, thousands of products, and hundreds of campaigns, channels, customers, and rules of the game, are more complex than what the human mind can process and solve - a barrier to progress. To address this, Celonis customers can leverage approximately 100,000 freely available applications in the Python community.

These packages combine process models with machine learning to help improve operations and understand the meaning behind them. The Machine Learning Workbench - currently in beta - was developed specifically for Jupyter notebooks and integrates Python predictive models directly into the Celonis platform.

Celonis also announced the acquisition of Banyas.it in early April, overcoming the complexity barrier of real-time process mining on SAP. In the case of high-volume or high-frequency processes, real-time process monitoring and alerting is critical to achieving targeted business outcomes.

Companies can see their SAP processes in real time and, above all, can take kick-off measures to avoid downstream problems or take advantage of opportunities. This brings SAP's existing customer closer to an ERP Digital Twin - much like Professor Sabina Jeschke is aiming for a Digital Twin for the railroads.

Celonis Action Engine is a solution for avoiding bottlenecks in the organizational and process structure: The system makes recommendations for action. In this way, people can implement the changes immediately or appropriate automation tools can react.

Based on business rules or machine learning algorithms, companies can create so-called skills in the Action Engine. These provide specific recommendations on process problems - before they occur.

Silo structures repeatedly cause business processes in companies to be out of sync. All employees must be involved in order to improve processes in the long term: It is essential to make them aware of the value of process optimization measures.

Celonis' Open Application Framework enables organizations to develop personalized, application-specific, intelligent and operational applications.

The framework includes best-practice templates and a development environment. There, apps can be created that combine transactional and analytic functions, allowing a user to intervene immediately while evaluating the analysis.

Transformation and Disruption

"Transformation strategies are now a fundamental part of any organization's operations, as no company can avoid adapting to the latest industry and technology trends"

comments Alexander Rinke, co-founder and co-CEO of Celonis.

"However, they should be based on concrete insights derived from processes that actually take place in a company. Our study shows that too many companies rush into expensive initiatives that are sometimes not even necessary and don't really add any value."

Results of a study published by Celonis show that many companies start transformation initiatives without knowing what the first steps or goals would be:

58 percent of executives surveyed say they don't know where to start when developing their transformation strategy. The study, commissioned by Celonis, examines how companies approach transformation programs, as well as the perspectives of executives and their respective employees.

The global survey of 1000 C-level executives and 1000 business analysts found that many companies have already spent significant resources on poorly planned business transformation initiatives.

In fact, more than half (56 percent) of executives in Germany believe their transformation programs have been a waste of time. This is despite the fact that more than a third (36 percent) of them say they have already spent more than 500,000 euros on transformation initiatives in the past twelve months.

As a result, they run the risk of incurring enormous costs with no discernible gain. It's no wonder that transformation initiatives often miss the mark: German executives often make far too little use of their employees' expertise. The survey shows a clear discrepancy between the management team that sets up transformation strategies and the employees who execute them:

Another white paper from Scheer logically states: In order to achieve the goals set, complete transparency of the processes is necessary, not only at the conceptual level, but especially also at the executive level.

In addition to classic, but often unspecific key figure considerations, topics such as the recognition of process variants including their significance as well as the identification of labor-intensive sub-steps taking into account the resource situation are relevant there.

Increasingly, the technique of process mining is being used. Due to its origins in Big Data analysis, this is able to deal with mass data efficiently and quickly.

Although process mining as a technique has been available for some time, it is digitization that now provides the necessary data on customer activities and internal processing to provide a complete end-to-end process view at the necessary granularity and speed.

Process mining is now no longer just a tool for classic process improvement, but is also used as an integrative tool of innovative operating models in the operational control of technical and human resources. (end of quote)

Digital Twins of an Organization

Professor Wil van der Aalst is a Dutch computer scientist who works on process management and is considered the father of process mining. He has held an Alexander von Humboldt Professorship at RWTH Aachen University since January 1, 2018. Previously, he was a professor at Eindhoven University of Technology.

At a Celonis event, he spoke about Process Mining: Past, Present and Future. One of the future tasks is DTO - Digital Twins of an Organization. According to van der Aalst, the scene is only at the beginning of process mining.

This also raises the question for Professor Scheer: Is the Digital Twin also a concept for business management processes?

"The Digital Twin is originally a concept for the digital mapping of technical systems. In the context of companies and business processes, the Digital Twin is a digital image of the company. This used to be called 'corporate memory'.

From there, the Digital Twin is also a concept for business management processes, as the company is represented as a digital model and different scenarios can be planned, simulated and measured."

Thus, especially in the agile environment, there are advantages in reacting to changing competitive conditions; predictive analytics also provides the opportunity to anticipate this. However, enterprise models contain more information than pure process representations.

It is a representation of the business and IT architecture from strategy to technology. Professor Sabina Jeschke, Board Member for Digitization and Technology at Deutsche Bahn, is absolutely right in her call for high-performance computers and a Digital Twin to meet the future challenges of the railroad.