Host Intro (Gene Kim)

Recently, I had the opportunity to attend Data Day, a fabulously inspiring event held by the data group at Daimler Truck North America, which is one of the world's largest commercial vehicle manufacturers, with over 40 production sites around the globe and more than 100,000 employees, and publicly traded.

It was so amazing and inspirational to see how this group of people are attempting to unlock the value of the immense amount of data that's buried within the company's systems and using it to create business advantage around some of their most strategic areas of strategy and operations.

I was so dazzled by what I saw, I asked Doug Murphy, who leads this program as their IT Manager of AI and Analytics, if he could present on what he's been working on. I'm so delighted that he said yes, and that he'll be co-presenting with Raquel Kusters, Product Manager of Data Analytics and AI. Here's Raquel and Doug.

Opening Video

We don't build trucks. We don't make engines. Don't make transmissions. Don't develop technology. And we don't manufacture buses or chassis for parts.

We create possibility. And we create.

We stock shelves, build bridges, hang cable, pave roads, move people, haul logs, hogs, pipe, grain, bricks, dirt, cattle, fuel, and anything made, grown, bought. We fuel economies and imagination.

We inspire the next generation of teachers and doctors, parents and children, explorers and dreamers. And we protect the next generation of teachers, doctors, parents, children, explorers, and dreamers.

We build futures and communities. We support towns, support businesses, support families and individuals. We support dreams, hopes, aspirations, wants, desires, big and small.

We've been doing it for 80 years in the communities where we build and where we live, serving hundreds of millions of people in North America and making lives better for the hundreds of thousands we, our dealers, our suppliers, and our customers employ, helping keep people, helping keep them housed, and helping them stay secure.

We create tomorrow. A tomorrow that's cleaner, that's healthier, that's safer and smarter. A tomorrow we all want to see. Tomorrow we all want our children and their children to see. Without us, the world would be a much different place.

Because we don't build trucks. We move the world.

Doug Murphy

I hope you got the message from that video. We do, in fact, build trucks, and we do it quite well. In North America, our Freightliner brand dominates the on-highway segment for both large and small fleets. Next time you are on the interstate, start looking at the big semi trucks. There's an almost 50/50 chance the next one you see is a Freightliner. You may also see one of our Western Star trucks or a Thomas Built school bus out on the roads. To learn more about our company, go to DaimlerTruck.com.

Hello. I am Doug Murphy. I lead the Data Intelligence Hub, which is Daimler Truck North America's center of excellence for AI and data analytics.

Raquel Kusters

Hi, I'm Raquel Kusters, and I am a product manager within the Data Intelligence Hub, working on AI and analytics.

Doug Murphy

Our goal is to take Daimler Truck to the next level of leading our industry: from trucking solution provider to game changer, from market leader to market maker.

In my role, I work with and talk about data every day. The data alone cannot make the kind of transformation happen. Well-managed, well-governed, accessible data is necessary. That is the entry fee for the kind of transformation I'm talking about.

And we have a lot of data, but it's the experts in each business area who know the data and know the business and have the imagination to envision a new process or a new business model derived from data, creating intelligence out of that data.

My team has adopted this statement as our purpose: we create intelligence for those who keep the world moving.

So before I tell you how we are transforming our business, a little background on our organization. In 2020, we launched our center of excellence as consolidation of three BI analytics teams within IT. We set course on a single managed data architecture, and we started the foundation of our self-service data platform.

And we worked with our Connected Services Group, vehicle engineering, and many others to start the first analytic use cases using streaming telematic data from connected Freightliner Cascadias operating out in the real world. This laid the foundation for our IoT analytics data lake in the cloud.

Everything we have built toward our target data architecture was done with business value-add each step of the way. Each use case does its part to put foundational pieces in place. We track the ROI of each and every use case we develop and run to ensure we are providing measurable value for customers, for dealers, for us, from everything we do. Our current metrics project and report internally on compound annual growth rate and measurable ROI from data use cases.

Taking that approach into 2021, we developed a solution to forecast replacement parts demand in the new, extremely volatile supply chain environment we all found ourselves in. You see, in the trucking industry, uptime is the most important success factor. A truck that needs repair waiting for that part that's on backorder means that truck is not moving goods, and frequently the driver is idle along with it. This costs our customers big money every hour that truck is not moving.

So we built a model using repair history. It gives us insights about the likelihood of failure, part by part, and within specific populations of trucks. Then we used connected vehicle information to see how those trucks were distributed across the map. The newly prioritized parts are now ordered and distributed to our parts warehouses across the country in the amounts suggested by the model. So far, more than half of the parts picked by this model were actually purchased and installed on a truck that needed it, equating to thousands of trucks that didn't have to wait for a backordered part.

This business-value focus has additional benefits. I'm proud to say my team is highly networked across the organization, working side by side with subject matter experts on some of DTNA's biggest challenges.

In 2022, we developed more advanced use cases, precisely offering needed aftermarket parts for large customers and leveraging connected vehicle data to anticipate emerging product needs.

In the very near future, 2023 and beyond, we see the potential to accelerate this innovation and to noticeably disrupt the transportation industry. Now Raquel will take you on a deep dive into some of our very latest innovations.

Raquel Kusters

Daimler Truck North America has a unique position in the market where we have 40% market share. As well, our leadership team took a strategic decision to have our vehicles equipped with telematics devices, where we are able to collect and store customer data with their consent.

Now since that decision was made, we have hundreds of thousands of connected vehicles on the road today, billions of miles traveled, which is made up of thousands of fleets. This data has been fully anonymized, so there's no specific vehicle or customer data here.

Every five minutes, we receive a GPS coordinate of our vehicles moving across our transportation network. The visual here represents one year worth of data, specifically 2021. The total amount is interesting and gives you a grasp overall about what roads are most frequently traveled. However, there are additional insights we can unlock that lead to true business value. But first we had to get acquainted with the data, so let's start off with some examples.

We did reverse geocoding on the longitude and latitude to identify top cities by trip, most frequent start city and end city, also a list of cities passed during a trip. One of my favorite ones so far is the map around most traveled routes. Imagine looking at an air traffic control map for airplanes. Well, this is the ground traffic control for actual trucks.

We also evaluated city versus highway driving, city to city, complexity of the route by elevation change and straightness. We have an early model for classifying the stop by category of a load, unload, refuel, or if that vehicle is actually taking a break. As you can see, there are lots of ways we can evaluate a geospatial data set.

Now let's take this one step further and talk about a use case that led to true business value. After spending some time deriving all those interesting insights, as I just shared, it was time to consider how we could use this to deliver exceptional impact for Daimler Truck North America.

As you saw in our introductory video, we have new technology coming to the market that will transform our ecosystem. That technology is electric vehicles. To ensure successful adoption of these vehicles, we as an OEM know there are new dependencies for our customers that they really need help.

This leads to the new infrastructure required to support charging our vehicles. The Data Intelligence Hub helps support the effort of Project Juno, which is a joint venture with BlackRock and NextEra Energy.

The mission of this new venture is to build the public infrastructure we need to help successful adoption of electric-powered commercial vehicles. What we did was help identify the high-priority locations for public charging infrastructure, provide the greatest benefit for the commercial vehicle market.

Utilizing our telematics data and insights, we identified three prioritized regions that would deliver the greatest return on investment. By making this data-based approach, we were able to uncover routes that are strong candidates for, say, diesel-powered vehicles, but say less desirable for battery electric based off range restrictions and stop durations of vehicle.

This way, we are directing our efforts to the locations that will have the greatest impact for our customers and really help ease that transition from diesel fuel to battery electric. Now I'll hand it back to Doug to discuss more about some lessons learned from this process, as well as how we manage and prioritize our workload so we can provide more high-value solutions just like this one to our organization.

Doug Murphy

Disciplined businesses in tight competitive markets like ours cannot easily do this. Most of what we work on year in, year out needs to be done within a fixed amount of human resources. So this becomes not a resource problem and constantly asking for more and more people; it becomes a prioritization problem.

Along this journey, we've learned some lessons about what works and doesn't work in scaling AI and analytics in an organization like ours. We've encountered duplicate data science projects happening company-wide. Rapid business innovation is often underleveraged with no on-ramp to sustainability. Missing opportunities to focus on the highest-value solutions because we are not aligned on prioritization. We have a lot of great talent scattered around the organization, but not leveraging them to scale appropriately at a company-wide level. And we also learned that we could be leveraging data assets, creating completely new revenue opportunities, opportunity we're not currently taking advantage of.

So how do we get aligned on priority? I know what I'm working on, but I may not know what data topics others are working on in the organization, and they probably don't know everything I'm working on, Raquel's working on. And we don't always know when someone is about to come to our team with a large production request that we're not equipped to work on.

So transparency is a big start, but we need a method of prioritization that is fair and aligned with DTNA's business objectives. We need a data-based decision-making process for prioritizing data use cases.

What I'm showing here is a representation of what we call our use case innovation funnel, and we've been refining it over the past three years. It is still not perfect, but I believe it is the best we have for addressing this topic. New ideas come in at the top of the funnel. The greatest width here indicates the largest number of use cases. Not all of them make it to the next stage.

We've developed a scoring system that puts measurable business value first. Only those that continuously remain as the highest scored in each stage will make it all the way through. We check for things like data availability, data quality, and platform or technology capability. Initially, we theorize about business value, and then in later stages develop an articulated business case.

In each subsequent stage through the funnel, the number of active use cases gets smaller while our commitment level to each one becomes greater. We scale the process based on effort level. For a 40-hour effort, it can be a single conversation and we're off to work. A 200-hour effort gets a more rigorous treatment. Our very biggest use cases require a fully articulated business case validated by our financial controlling.

The intent is that we will soon use this process collectively for all of the Data Days. Business data workers bring in use cases at their inception, and we'll go through the scoring process together. If the use case is prioritized, we can align planning and resource timing across our teams and any other needed expertise.

We'll all have transparency: what are the highest business-value use cases being developed across the organization. Now we'll also have to accept when the scoring is low, and together we must prioritize the other highly scored work. This doesn't work if people go rogue and around the process, back to developing within their own silos. So I'm asking my colleagues here, if they see problems in the methodology, not to give up and help us make it better so that everyone is operating within this new model.

I believe this is the best way to channel DTNA's fast-growing pool of analytics talent into the use cases that we have so that we have the highest value to our customers. So as we close, I want to ask the audience, first of all, if you've got any feedback on what you see, I'd love to hear it, and we'd really love to hear your ideas on innovative use cases we could take on based on the data and analytics that we showed you today.

Thank you all, and really appreciate the opportunity to share what we're doing here within DTNA.