Main content

BMW drives process mining through its business with Celonis

Derek du Preez Profile picture for user ddpreez April 24, 2024
What started as looking at individual processes at BMW back in 2016, using Celonis’ process mining technology, is now expanding to interdependent processes and interdependent suppliers to achieve more value.

BMW 4Series Convertible

BMW Group, one of the world’s leading manufacturers of premium cars and motorcycles, has expanded its use of Celonis’ process mining platform to further break down operational silos across its organization - as well as reduce friction with its suppliers. 

BMW has been using Celonis since 2016, where it initially focused on mining individual processes in different areas of its operations, in an attempt to extract value and improve efficiency. However, following the success of its Center of Excellence and seeing the support for adoption, it is now expanding its use to the vendor’s latest object centric technology, which will enable BMW to map the relationship between different processes, to identify pain points and seek out further efficiency gains. 

Further to this, BMW is also looking to make use of Celonis’ generative AI features, which aim to democratize process mining for buyers, allowing business users to ask questions of the platform, in order to gain insights about their processes, without needing to be an engineer or data analyst. 

There are currently over 200 Celonis use cases across business units at BMW, as well as over 450 Celonis data models and 1,100 process automations. Dr. Patrick Lechner, Head of Process Mining at BMW, explained to diginomica that the early ambitions in the adoption of Celonis back in 2016 were to make BMW Group a more data-driven company. 

The organization started by looking at just two processes: purchasing and production. However, this quickly expanded after some early wins. Dr. Lechner said: 

We started very small, just a colleague and myself, back then in 2016. But we soon realized that if we want to use this in a global way, and we saw the potential quite fast, then we needed a central Center of Excellence (CoE). 

Otherwise it’s always stuck to one business unit. IT at BMW is at the center and is global, which is why we established the CoE there. At the moment we have people in Munich and Porto working for the CoE. 

Once we saw scaling was possible, we decided to support business units for different prime processes. And from there, we scaled the CoE further, introducing Centers of Competence in the main business areas, so that we have experts in those areas providing support. They identify processes that may be relevant, where there are pain points, and suggest where process mining could be used. 

Adopting at speed

One of the first Celonis use cases in production was in the BMW Munich paint shop, back in 2017. Dr. Lechner said that in many cases his colleagues in the paint shop had been working in that area for the last 20 to 30 yeaars, which meant that whilst they were open to exploring the capabilities of process mining, the understanding of where it could be applied largely came from them. He said: 

They were very interested in how a new technology could support them. We looked at the different process steps and could see: how easy it was to paint a certain color; how long it took; the energy needs we have, etc. And we supported them to optimize these processes, to reduce the times, reduce the energy costs, and so on. 

Establishing the benefits in this one paint shop meant that the technology could then be applied to other BMW plants that were using similar IT systems, as well as other functions such as assembly, body shops, and other parts of the production process. 

One challenge that is common when it comes to process mining is getting the business itself to acknowledge that the data being delivered is noteworthy and that processes could in fact be improved (as is often the case with technology adoption, culture and change management are actually the biggest barriers to results). However, how BMW approached this was to let its business colleagues guide where they thought problems could exist, initially, in order to get buy-in to the tool. Dr Lechner said: 

Business normally know its processes best, so it always makes sense when the initiative comes from the business. It’s also sometimes useful to bring business units together, so that we get out of those silos. 

But most of the context comes from the business, because they want to use the tool - and it’s our job to make sure the tool is well known within the company. This is very important. If you have a company the size of BMW, you have to make the tool well known. 

The ideas normally come from the business though, where they can see pain points or potential, and then we can support and help them implement it. We do consulting in the CoE, support them on implementation, and also look after the platform. 

The BMW CoE includes data engineers who connect the IT systems to Celonis and make the interfaces work. It also has analysts, who then help the business units create dashboards, and as Dr. Lechner notes, there are consults who work to really understand the business needs. He added: 

Ideally, you would be able to combine these roles in one person, but this is always a big challenge. And then you need people who are good at operations, who support the running of the platform, as well as the security topics and making sure we are compliant. So there’s a range of roles we combine in the CoE. 

Looking beyond single processes

Since the BMW/Celonis partnership began, Celonis has since expanded the capabilities of its platform. As noted on diginomica, the vendor is now focusing its efforts on ‘object centric process mining’, which allows buyers to map beyond single processes.The best way to think about it is that whilst process mining is very effective in following a single object through an organization (such as an ‘order), it fails to take into account other objects that impact its progress (such as production, shipments, procurement etc). 

Understanding how all of these ‘objects’ interact with each other, in real time, can allow an organization to fine-tune operations in a way that previously would have required manual workarounds and some inefficient guesswork. OCPM - which underpins Celonis’ new Process Sphere technology - maps how all these moving parts intertwine. 

This is now what BMW is pursuing internally. As Dr. Lechner explained: 

Yes, we are currently doing that [moving to object centric]. We have the first use cases in purchasing, customer support and production. We definitely see big potential in it. But it is a different approach, you have to build those objects first, you have to model it in a new way where the objects are interacting. 

If you want to do object centric with scale, you have to set up the right structures, and we are currently working to set up those structures at BMW - but in isolated use cases we are already using it. 

However, despite the effort required, there are benefits to be gained. He added: 

This will help us really understand how the different processes interact. So far, it’s always been following one process through and you can’t see how it’s interacting with another process. 

For example, how is the test drive process interacting with the customer? How is the production then coming in? How is the financing coming in? All these different parts of BMW we can finally see how they interact. Hopefully we can get new insights and optimize the process globally. I think for the next step, object centric can really help us. 

In addition to this, the BMW Group's Center of Excellence for Process Mining and Celonis' development team also agreed to collaborate more closely on joint process intelligence innovations, such as connecting suppliers in cross-company processes more efficiently. This will be done by automating the transfer of data (e.g. via CatenaX)  and mapping customer journeys. Dr. Lechner said: 

This is the next step, because we are moving away from the silos using object centric process mining. But also, we sometimes have to optimize our processes together with suppliers. If you look at the supply chain, and you want to understand why a supplier is not able to produce in time, I think it will help to have this transparency with suppliers. 

But this is also a big challenge, because we have to exchange data. It’s easier if it’s just in your company. But we are in trials to see how we can do analysis together with suppliers. And it looks quite promising, what we have seen so far. But of course you need use cases where both can benefit from this, the supplier and BMW. 

Further democratizing process mining

Key to BMW’s plans for process mining is further expanding its use to business users, so that the technology doesn’t always have to rely on data analysts and engineers to deliver results. Dr. Lechner would like to see regular business users asking questions of the Celonis platform when they think they might recognize a process pain point - and the key to this is going to be Celonis’ generative AI developments. Dr. Lechner said:

We really want everyone to be able to make those data driven process decisions. And for this we need tools that are very easy to use. I think the user friendliness of Celonis, the ability to build dashboards and so on, is really good. 

However, if we want to go to the next level, we need people who don’t understand process mining to interact with the tool, but may just ask questions. And for this, we are going to be evaluating the generative AI possibilities with Celonis. 

I think this will really help to roll it out to even more users. Our target is that everybody can be able to ask those process questions, in a very easy way. And then get feedback from Celonis, for the analysis. I think this will be crucial in democratizing it completely. This is the target, they don’t need to know it’s a process, they just have a business question that is process-related, and then just get the right answer. 

A grey colored placeholder image