By Christine LaFave Grace, Plant Services managing editor
Zooming by at 70 mph, most of us would barely give a second thought to a semi broken down on the side of the highway. But Navistar, a leading manufacturer of commercial trucks, buses, engines and defense vehicles, thinks plenty about this frequent sight, because it knows exactly how expensive these breakdowns are.
"Every time a truck goes down, it costs the operator of the truck $750, not including repairs," said Dan Pikelny, Navistar's VP and chief analytics officer, at the 2017 Smart Industry conference. For a company with the motto "Uptime is our mission," the status quo in terms of trying to avoid breakdowns and addressing them when they occurred wasn't cutting it. The same old maintenance strategies would produce the same old results, Navistar knew.
"Hindsight is 20/20," Pikelny said. "We want foresight to be 20/20." So to promote a more predictive approach to servicing its vehicles, Navistar created OnCommand Connection, an open-architecture advanced remote diagnostics system.
Using the OnCommand system, highly connected trucks are able to send data to and receive data from the cloud—everything from geospatial data and road-conditions information to diagnostics data. Vehicle-to-vehicle communications capabilities and the capacity for over-the-air programming of engines (for Navistar and Cummins engine models) mean that operators get more-comprehensive, real-time information about the status of their trucks and the road ahead and can better identify and troubleshoot emerging issues.
"If you think about Facebook, this is like the social network of trucking," Pikelny said. But far from acting as merely a sophisticated crowd-sourced traffic report, OnCommand allows for aggregation and historical analysis of vehicle data to help identify driving patterns, statistically hazardous areas, etc.—information that can save operators money by helping them steer away from trouble.
Real-time feedback
"Where are trucks cruising? Where is harsh braking required?" Pikelny offered. One of the big factors in the cost of operating a truck is road quality, so it's valuable to have that data always on hand, he said. Which trucks are regularly running on lousy roads, and what does that mean for their service life? Using all of this telematics data, combined with data about a vehicle's make/model, engine, model year and more, individual vehicle risk models can be developed, Pikelny said. Further, failure prediction modeling can predict the risk of failure for a specific issue based on relevant data sets. Armed with this information, customers can make smarter calls about exactly when to pull a truck off the road—i.e., before a breakdown occurs or performance/safety is affected but at a time that minimizes the financial impact of a truck being out of service. For drivers themselves, ease of use is enhanced, too—a truck can get the latest software update over the airwaves when the truck is stationary rather than having to go to a dealer, Pikelny said.
Moreover, from a vehicle design perspective, this wealth of data from the more than 300,000 vehicles that OnCommand Connection currently supports offers powerful guidance for the building of future vehicle models. "Warranty is a huge cost for many manufacturers, including us…data helps us design extended warranties," Pikelny said. Also, that the fact that the OnCommand Connection system provides integrated support through Navistar dealers and parts and service providers can be a powerful consideration when it comes time for customers to make purchase decisions.
The result for OnCommand Connection: so far, a 30% reduction in maintenance costs for Navistar vehicles, Pikelny said, and a significant reduction in roadside failures.
Pikelny shared words of wisdom for fellow OEMs looking to close their own loop on product design and deliver stronger value for customers. Navistar also discovered disconnects between its expectations for digital processes and the reality of them. Specifically, with the launch of OnCommand Connection, Navistar's data team expected to spend a moderate amount of time on data preparation, the majority of its time on analytics, and a small amount of time on data integration, Pikelny said. The reality was more or less the opposite: a small amount of time on the analytics themselves, with most of their time spent on data prep, integration and helping stakeholders best make use of the data. After all, no analytics team can force the business's customer—or another department within the business itself—to make smarter decisions using the information at its disposal.
