T18 panelists discuss managing telematics data flows

 From left to right: Eric Whitton, Mark Ehrlich, Gerry Mead, and Roni Taylor discuss connected trailers. ( Photo: FreightWaves )

From left to right: Eric Whitton, Mark Ehrlich, Gerry Mead, and Roni Taylor discuss connected trailers. (Photo: FreightWaves)

Two of the last panels at Transparency18 on Wednesday afternoon were on telematics: the first conversation touched on freight visibility, vehicle diagnostics, and driver safety, while the second panel focused on connected trailer technology. A common theme during both discussions involved ways of managing the enormous data flows from telematics devices—how much data is actually needed to solve a given problem, who owns it, and how best to leverage it in a way that respects intellectual property and privacy.

Dean Croke, FreightWaves’ Chief Analytics Officer, moderated the first panel, whose participants were Sandeep Kar, the Chief Strategy Officer for Fleet Complete, Bill Kozek, the President of North America Trucks and Parts at Navistar, and Jason Penkethman, the Chief Product Officer at Spireon. Each panelist began with some introductory remarks about their own vision of the telematics industry.

Jason Penkethman of Spireon began, “The maturity of the telematics industry is happening at the same time we’re seeing other technologies mature, like machine learning; applying that to telematics is a huge benefit. And everything you heard this morning around blockchain is another interesting area where you can apply telematics. It’s exciting because these converging technologies that can accelerate adoption and the transformation of the enterprise.”

Bill Kozek said Navistar’s interest in telematics was directed at an all-important KPI for fleet owners: “The goal is ultimately to develop a vehicle with zero unscheduled downtime.”

Fleet Complete’s Sandeep Kar initially sketched out a more philosophical position, saying that for many decades, transportation companies have purchased assets and then had to figure out how to use them, but that in the near future, the approach carriers will take might sound more like “Here are the services the truck provides, what can it do for me?” Kar seemed to be referring to the concept of Transportation-as-a-Service (TaaS), where fleets of autonomous, electric, and connected trucks owned by OEMs are hailed and matched with loads by transportation companies on an as-needed basis.

One of the most interesting concepts that came up in the discussion was the prospect of edge computing, which is a strategy for optimizing cloud computing. The term comes from the idea of locating computing resources at the ‘edge’ of the internet, where the cloud interacts with the physical world. In practice, this means that more data retention, analytics, and recommendations would happen on the telematics device itself, on the vehicle. Instead of receiving a steady, high-volume stream of raw data from the device, an OEM or fleet owner would only receive actionable insights generated by algorithms: they would only get the data that made a difference.

Penkethman seemed to think that edge computing has yet to come into its own. When asked about remote vehicle diagnostics, Penkethman commented, “There are different degrees of complexity… you start with miles, then idle time, then weighted diagnostics, and then on the edge side, I think we still have a ways to go, I like the concept but it’s a ways out.”

Kozek was a little more willing to speculate about the future potential of powerful telematics devices, envisioning “over-the-air firmware upgrades to optimize performance for different topographies” where engines and drive trains could be re-specc’d in a mountainous or flat region on the fly with geofencing technology. 

One of the questions regarding the management of data flows was “How real time is real time?” The respondents specified that different levels of visibility were required depending on the transportation mode and type of freight. “It depends on the type of freight and the type of asset delivering it. For a container that’s on the ocean or a train line, you probably don’t need it every second—it needs to be applied to the correct use case. But a lot can happen in two minutes, from a safety or a security perspective. People who do cross-border want much more granular data, for instance,” said Penkethman.

The second panel, on connected trailer technology, was moderated by Roni Taylor, VP of Industry Relations at Spireon, and included Gerry Mead, Executive Director of Innovation at Phillips Industries, Mark Ehrlich, Director of the Van Product Line at Wabash National, and Eric Whitton, the SVP of Technology at Covenant Transport. 

Roni Taylor began by describing the present-day data management problem: “Trailer management goes back to about 1999—that’s when people first started putting telematics devices on trailers. They thought it was fantastic that they could get a location position at 6 AM and then 6 PM, but they didn’t really know what was happening in between. Now what we’re seeing is this—we’re being inundated with so much potential data.”

Eric Whitton of Covenant was asked to define a ‘smart trailer’ and replied, “as a consumer of the data, I would paint with a very large brush and say that a device would proactively communicate relevant information to the parties that need it when they need it.”

“A smart trailer and a connected trailer are really two different things,” said Mead. “A smart trailer can call you and talk to you and tell you its issues. A smart trailer can’t wait for a user to request its data, a smart trailer must bring the actionable data to you before you request it, instead of waiting or it waiting for you to queue it.”

Mead continued: “It has to be edge-computed and the consumer only wants data that can be acted upon. It’s all about who can take that data and make it actionable and meaningful, and that will stop the data overload at that point, and you must do that at the edge or at the cloud and transmit it almost instantaneously to the fleet.”

Mark Ehrlich of Wabash took up the OEM’s perspective: “When we look at action-based, who is the actor? Is it the fleet, or from the OEM perspective? Ultimately we’re creating value from this data set with the goal of making the fleet first and foremost. There are thousands of variables… the trailer is interesting because it’s a rich data source because it can indicate things about the commodity being hauled…the inside of the tractor looks pretty much the same all the time, but the trailer doesn’t.”