Use Big Data and Predictive Analytics to Drive Innovation

This content relates to : DIGITAL TRANSFORMATION

Kirk Andreae

CEO, Clarke Power Services 

I’m Kirk Andreae, CEO of Clarke Power Services. What we’re going to talk about now is using data to drive innovation. So when our business, and we’re talking about our Vehicare business right now which is a fleet management company, which works on the assets of a fleet. So it can be a distributor of beverage products. It could be a long-haul trucking company taking assets from an Amazon warehouse to an Amazon distribution facility. So, we’re talking multiple types of assets and in our business, what we do is we focus on the preventive maintenance aspect of any of those assets that the customer has. That is our introduction to the customer. What that does is that gives us a fleet list or a list of those assets and the attributes that those assets have. The more of that data that you gather, the more information you have on how those assets operate. Especially if you are seeing those assets on a preventive maintenance cycle at any one of those customer’s facilities.    

So basically we operate on three axes. One is the product axes, and that could be a trailer, it could be a truck, it could be a forklift, it could be a conveyor belt. We also then look at the attributes that are within a market. And these are also customer defined market attributes which is how they utilized the product. And so some dissimilar industries, like what would you think of as an Amazon distribution warehouse versus a beverage facility actually operate on the same equipment. So you can combine those markets from their attributes into a single market because the definition of that market is the products that they use. So that then gives you expertise across multiple markets. And then you can look at the geography. If I have a customer in a specific city, say Nashville, and that customer also has a location in Indianapolis. We can use the data that we’ve trapped in Nashville to predict out what the costs will be in Indianapolis and go to the customer with a warm lead because we already work internally in their business. We can go to Indianapolis and say “Here’s where your costs should be” And we trap that through our service orders, which is our point-of-service information, which is filled in by our technicians and our service advisers. And then we also utilize a telematics product that we hook in to each one of the assets.    

It’s a multi-use product, so it can connect to a towable, which would be like a portable chain set being pulled behind a truck or a car or it could be a trailer or it could be a truck itself. So a class A truck, class 6 truck, or an automobile and it transmits thousands of points of data every second. And so we utilize that data to predict where in the lifecycle, not only on the preventive maintenance but also on the predictive maintenance side, what components will fail? It’s all based on history because we’ve trapped all of this information before. So the longer we work on the assets, the more that we understand where the failure points will be based on a certain set of operations and what year that asset was purchased. So you do it basically on an age band model. We’ll look at where is the use of the asset where is it in its useful life. And then you begin to predict out the costs. You need the telematics data to do this, because every asset has its own life and so you have to be able to predict out where that asset is in its useful life. That then drives you it not only into different products and different supply chains, but it defines the different markets that you may not have thought were some similar. So, the data triangulate all that information, tells us where they are in the life-cycle and then we can predict out the cost and give a defined outcome to the customer within a set of parameters that are always defined by the customer.    

The key to the data, though is while we may view something like is simple, our customer may not understand what we’re discussing. So a golden rule of the data is you have to present it in a way that your customer can understand it and interpret. It has to be meaningful to them if you’re going to actually be able to sell that as a product or enhance the product that you are using. 

Kirk Andreae 

CEO, Clarke Power Services 

https://www.clarkepowerservices.com/