Besides linking real-time vehicle and internal repair data to visualize all this data in new ways, Collective Data can also bring in external spreadsheets and system data into the fold to create tables. A diagnostic-minded leader could theoretically take this data and ask AI to generate a fishbone diagram to troubleshoot vexing shop issues.

It’s not uncommon for these tools to directly help troubleshoot issues in the bay and the back office. Similar to Fleetio’s AI Service Advisor, Motive Maintenance launched in August and bridges inspection and vehicle/asset health data with the repair workflow in the shop.

“A lot of deep dives can be done, because all of the data that we are now building with Motive Maintenance also sits in Motive Analytics,” said Sri Teja Kolluri, staff product manager at Motive. “That allows customers to diagnose some of these bigger questions, like what parts are being replaced the most? Then they actually see the fault codes from these vehicles and come up with a predictive analysis.”

Meanwhile, the tech can receive important time-stamped data—such as DTCs, GPS info, and engine RPM—contextual data that fosters a more informed diagnosis.

“When they’re opening the work order, they would view all that information, including when the DTC was first observed and when it last occurred,” he said. “They don’t even have to open a different diagnostic screen for that. You can do that entire investigation within Motive Maintenance.”

Every trouble code tells a story

The investigation can zoom out and zoom in as the maintenance team needs to address immediate and long-term issues.

Let’s say a fleet put off resolving DTC P2463 (DPF soot accumulation) on one truck too long and had to replace a head gasket. The manager, wanting to prevent more costly downtime, would likely want to know how prevalent this code is and the cost of fixing now versus waiting.

Under normal conditions, it would take an expert familiar with that code several days to manually figure that out, offered Sean McGee, VP of product and engineering at Samsara. And “for many fleets, it would be impossible to really do that right at scale,” he said.

With AI, it can be instantaneous. “The AI can scan all the historical patterns and then give you a recommendation based on this fault code, [answering] what’s likely to happen next?” he said.

McGee demonstrated this to Fleet Maintenance using a new Samsara Connected Asset Maintenance feature called Enhanced Fault Code Insights. He clicked on a unit that had a fault code, and the description popped up. Then, because the asset was a 2024 Freightliner Cascadia that has a Detroit Diesel DD13 engine, it provided more context based on aggregated data from 100,000 other similar assets in the Samsara network that had that fault code.

The system showed a sensor calibration would likely fix the issue. But then if you wait, it predicted 23% of the time a new code would pop up within 518 miles. And from there, nearly half the time the issue escalates to throw another code, which, if unaddressed, would lead to an expensive roadside repair.

For a scenario like this to work, scale is vital.

“To tell a story about this fault code, you got to have a lot of these,” noted McGee, who said Samsara has accumulated trillions of data points from millions of vehicles in the last decade.

Enhance, not replace

There’s no denying that AI will disrupt routines and displace workers. If you spend your days manually entering invoices, you’re out of luck. AI can scan and contextualize those in seconds.

But for shop managers, this can “eliminate the grind,” Samsara’s McGee offered. He pointed to the company’s AI invoicing scanner that directly addressed a major time-suck for shop managers.

“On any given day, they’re inundated with invoices from vendors,” McGee said, explaining they have to read them, key in the info, check what work was done and parts were used, categorize them, and check if the fleet was overcharged by comparing the billed hours to the SRT. He said “millions of hours are being wasted” across the industry by doing these tasks manually.

Samsara’s AI invoice scanner can do this instantly, which has freed up more than 100,000 hours so far, McGee said.

Trimble’s TMT platform has integrated Corcentric to do similar tasks, populating invoices and repair orders, assigning VMRS codes to help managers analyze data.

Brian Mulshine, the former senior director of TMT product management for Trimble, explained this granular view can “get it down to what it costs you per mile to operate that truck by system group, [such as] tires, air conditioning, brakes. And now you could look at where you’ve got to focus training and cost improvements and what’s driving those.”

But will AI evolve to a point where it takes that fleet manager’s job as well?