FP&A Talks Series: How AI Powers a 185-year Old Transport Company

FP&A Talks Series: How AI Powers a 185-year Old Transport Company

By Anders Liu-Lindberg, Co-founder, COO and CMO of the Business Partnering Institute

 

FP&A Talks is a collaboration between FP&A Trends Group and Anders Liu-Lindberg.

Imagine you’re a train dispatcher tasked with making hundreds of trains run smoothly every day. They must go through an interconnected network where trains run in a sequence and can’t just pass each other along the route. Suddenly a delay occurs on one of the trains because a door won’t close properly. What do you do?
 
Your decision can impact the commute of thousands of people that day and potentially cost your company millions in fines. It’s like the butterfly effect where a single train can cause mass disruptions throughout the network if not handled well.

It sounds like a tough dilemma yet it’s a situation that the Deutsche Bahn Regio train dispatchers face every day. Understanding what’s the best course of action involves estimating the effect of millions of data points in a matter of minutes. Clearly this is impossible for any human brain but for Artificial Intelligence (AI) it can be done in a matter of seconds.

This is just one of the uses that FP&A at Deutsche Bahn Regio is currently investigating to use AI for and the potential is massive as Tanja Schlesinger, Vice President of Business Intelligence OneSource at Deutsche Bahn Regio, puts it.

“When I look back at my own past with data, I wouldn’t have thought that anything was possible just five years ago – our data was not only found in old legacy systems but it was inconsistent – totally inconsistent…at the end of your journey as you find a more solid basis in data and information there are use cases just pouring down the tree-like ripe apples…when have you ever had the opportunity to work with that?” 

Yes, we’ve been talking about AI for years and now we start to see the use cases where the real value is added!

How FP&A pushes the transformation at Deutsche Bahn Regio

A company like Deutsche Bahn Regio is under heavy scrutiny and must comply with high-quality standards to maintain its business. It used to have 100% of the market but today the market share is down to 2/3s. Further to this, failing to comply with the quality standards will lead to high penalties to be paid to the public hand.

To maintain a high-quality standard or even make further improvements a different way of planning and looking at data is needed. When you know what impacts the company finances in terms of operational metrics it would appear obvious to make your planning driver-based. However, as always there are complications as Tanja, who’s also responsible for Business Intelligence explains it.

“ …for achieving a solid performance to our passengers we currently need more transparency on our quality levels because most of our quality levels include a tight measurement regime and any failure of contractually agreed thresholds is going to be heavily penalized…on the other hand our insights and access to transparency [into these measurements] is currently hidden in numerous old legacy systems and as we aim at value driver-based planning and analysis we need to get a grip on those legacy systems.”

Even when you get a grip on the legacy systems and feel you have the data challenge under control you still need to be able to translate business requirements to the data scientist (as FP&A professionals normally don’t have the needed programming skills). It’s no longer enough simply knowing how to translate the outcomes of the model to the business stakeholders. No, FP&A professionals must become multilinguistic translators that can speak both business and data. Tanja further outlines this challenge.

“The typical controller or FP&A staff has no programming skills and no experience in explaining a scripter what to do. So as we learn how to become a valid business partner we now have the challenge of becoming a valid business analyst and a scripter for a programmer to transport what we know into mathematical structures and into algorithms”

FP&A at Deutsche Bahn Regio has taken on this challenge though and is now pushing the digital revolution within the company. You might ask why is this a task for FP&A? Tanja tells us.

“…it’s weird to say that’s an FP&A topic [helping the train dispatcher make his decisions] but at the end I know that he’s not in the position with no tools at all to influence well something that he’s in charge of. For such a small thing we talk about within an hour 20 million data set. That’s out of scope for any human being!”

I’m sure none of us would want to take on that task of the train dispatcher but now with the help of AI he’s getting an autopilot that’ll help him make those decisions much faster and with higher quality. This will greatly limit the penalties levied on Deutsche Bahn Regio and improve the customer experience for the train passengers.

Don’t wait for others to take the lead on AI

FP&A is uniquely positioned to take the lead on data and AI because we span across the entire company and can gain access to all relevant systems (financial as operational) to bring together the needed data sources that power AI tools and help us make better decisions. What used to be an almost impossible task of gathering data from disperse legacy systems is now easier than ever. Tanja puts the utmost wish in perspective for us.

“…your utmost wish would be to have a driver-based model where you have a sensitivity mark like 10% up or 1% down and you watch your EBIT on the other side and say <well that’s not worth all the effort you put into working on the business driver>. So, on your driver set to finally find something that gives you answer to if it makes sense to spend an hour, a day or a month working on that item.”

The ideas of how to make the improvements will still come from the business but it’s up to FP&A to translate them into use cases. That’s at least Tanja’s final thought.

“The ideas will still come from our business units on what and how to find new ways of dealing with uncertainty and learning from historical data and new data to be grabbed from outside the company. In the end, though how to translate it into a successful use case that’s the role of FP&A”

This is indeed a new and exciting challenge for FP&A and at Deutsche Bahn Regio they’re showing the way and making the life of the train dispatcher easier is just the beginning!
 

The full text is available for registered users. Please register to view the rest of the article.
to view and submit comments