Government fleets do not need more information just for the sake of having it.

For many agencies, the big issue is what happens after fleet data starts flowing in. A dashboard may show where an asset is or when a driver safety event occurs, but someone still has to decide what that information means and what action should follow.

Several conversations with public sector fleet leaders and a technology provider during Samsara’s Go Beyond Public Sector 2026 event made it evident that fleet technology is only as useful as the process behind it.

The fleets seeing the most value are not treating the technoslogy as a fix on its own; they are using it to support decisions they already need to make.

For the City of New Orleans, the starting point was basic asset visibility.

The city used connected fleet technology across 41 departments, including New Orleans Emergency Medical Services (NOEMS) and the Equipment and Maintenance Division. According to a report on the city’s deployment, New Orleans is tracking 2,630 vehicles and assets through the platform.

Before the deployment, New Orleans had a fragmented view of its vehicles and assets. NOEMS had been using several methods to track vehicles and assets, including radio GPS, modem GPS, Apple AirTags, and magnetic boards.

The Equipment and Maintenance Division faced a similar challenge. EMD is responsible for managing approximately 2,700 assets, including police, fire, and maintenance vehicles. When Shelita White, operations chief of staff, and Kim DeLarge Jr., assistant chief administrative officer, came into fleet, the city’s asset records were spread across paper files and spreadsheets.

“We received just a stack of papers,” White said. “We literally had to scrub spreadsheets.”

White said one spreadsheet initially listed about 2,900 assets, but after the team reviewed and verified the inventory, the count was narrowed to about 2,600.

That lack of a centralized view affected more than daily convenience. Before the city could make stronger decisions about maintenance, utilization, or replacement, it needed more reliable information about the assets it already had.

With the system in place, the New Orleans fleet had a clearer view of where equipment was located and how it was being used. The city also used the technology to support maintenance planning, fuel oversight, and safety efforts across the fleet.

For New Orleans, the work showed that technology was most useful once the city understood which assets it needed to manage and how departments planned to use the information.”

New Orleans also used the technology to support emergency response. NOEMS uses telematics data to route vehicles more efficiently and keep ambulances and other vehicles current on maintenance. In that environment, reliable vehicle information is not just an administrative improvement, but can affect how quickly resources are assigned and whether vehicles are ready when they are needed.

The city also saw value in geofencing and fuel oversight, identifying situations where an asset may be in one place while a fuel card is used somewhere else. For a government fleet, that kind of exception reporting can help agencies ask better questions and strengthen accountability without relying only on manual review.

For agencies still working from paper records or incomplete asset data, that first step can be the hardest one. Before a fleet can use data to defend replacement decisions or improve maintenance planning, it has to trust the basic information in the system.

Implementation Requires More Than Installation

New Orleans did not roll out the technology by simply installing devices and expecting departments to use the system.

The city started with a pilot and held weekly meetings with individual departments. The goal was to understand what each department needed from the system, rather than assuming every user had the same priorities.

For the fleet maintenance team, the priority was asset and preventive maintenance tracking. Other departments came to the system with their own operational concerns, including safety and routing needs. Those early conversations helped the city frame the technology around each department’s work instead of presenting it as another requirement being placed on already busy teams.

White and DeLarge also described weekly internal systems meetings to track installation progress and determine how many additional assets could be brought into the system. Instead of requiring departments to bring vehicles in, the team moved toward on-site installations. That mattered because a busy department could easily delay installation for weeks or months if the process depended on bringing vehicles to a central location.

On-site installations helped remove a common obstacle by keeping departments from having to disrupt their schedules to bring vehicles in. Technology adoption is often framed around software, but implementation can stall when the process adds work for the people expected to use it. New Orleans reduced that friction by bringing the work to them.

The city also treated support as an ongoing responsibility, with White describing answering calls and emails, walking users through the system, and helping departments understand how to navigate the platform. That kind of support helped users move from uncertainty to confidence.

“The overload is just constantly having those weekly meetings, like I said, and encouraging them,” White said. “They know they can pick up the phone and call me anytime.”

This approach also helped the city build trust before pushing too much change at once. DeLarge said the department had to address basic needs first and give employees a reason to believe in the direction the fleet operation was moving.

A new system in a public sector fleet can reach well beyond the fleet office. The people expected to use it need to understand why the data is being collected and how it will be used. Without that clarity, technology can become another point of friction instead of a tool for improvement.

Data Can Support Safety, But Agencies Need a Process

Nebraska DOT’s experience showed what comes next, after a fleet already has years of operational data to work from.

Ben Merchant, assistant director of operations for the Nebraska DOT, said the agency had used AVL technology in heavy-duty plow trucks for years. The state’s 511 system includes a plow tracker that allows the public to view plow locations and images during winter weather events. Media outlets also use that information during storm coverage.

More recently, Nebraska DOT moved to Samsara as part of an effort to right-size its technology. Merchant said the agency has used cameras in snowplows to receive alerts for harsh braking, stop sign events, and crash detections. Those alerts can support coaching when needed. They can also provide useful context when a driver is involved in an incident.

“In several cases this year we’ve seen our employees vindicated of any wrongdoing in incidents that they were involved in,” Merchant said.

Nebraska DOT is also using fleet data to review winter operations after storms. Merchant said the agency can look back at reports to evaluate how a storm response worked and whether changes could improve future operations. The agency is also looking at utilization as it considers whether equipment is properly distributed across the state.

But Merchant also pointed to one of the biggest challenges facing public sector fleets: data volume.

Once an agency has more reports, alerts, and operational information, it has to decide how to use them. Merchant said Nebraska DOT now receives weekly information on speeding events, but that information has to be reviewed in a way that is fair and manageable. A report may point to a coaching need. It may point to a system issue. It may show that a posted speed limit in the system is incorrect.

“We had to have a system in place to process that information in a way that’s fair and manageable by our staff,” Merchant said. “Otherwise, all we’re going to do is sit around and stare at numbers.”

Fleets adding technology with limited staff still need a way to manage what the system produces. More information can support better decisions, but someone has to be able to review it, interpret it, and decide what happens next. Without that structure, data becomes another administrative burden.

For Merchant, the opportunity is not just collecting more information, it is finding better ways to process the information the agency already has and turn it into something supervisors can use.

AI Works Best When It Reduces Manual Work

Sean McGee, vice president of product and engineering at Samsara, said many public sector maintenance operations are still dealing with paper-based workflows. In shop visits, he said, he has seen work orders written out, entered into a system, printed again, and handled through multiple manual steps.

“There’s an incredible amount of inefficiency,” McGee said. “There’s a hundred little paper cuts along this process.”

McGee said AI should not be evaluated by how advanced it sounds, but by whether it removes a real operational burden.

For example, he described shops where staff members were manually entering invoices instead of spending time on the shop floor. In that case, the question is not whether AI sounds innovative. It is whether it helps experienced employees spend less time behind a computer and more time supporting maintenance work.

“I think it’s important not to just do AI for the sake of AI,” McGee said.

McGee also pointed to total cost of ownership as an area where better data could help fleets make more informed decisions. Rather than relying mainly on mileage or gut instinct, a fleet could look at whether a major repair still makes sense for a particular asset. That decision becomes more useful when the system can connect maintenance history with the broader cost profile of the vehicle.

For government fleets, that kind of decision support could matter as budgets remain tight and replacement cycles stay difficult. Still, the broader message is not that AI will solve fleet management, it is that AI may help organize information in a way that makes decisions easier to defend.

And as Merchant pointed out, the value is not simply in having more data. It is in being able to understand what that data means.

“It’s not just raw data that’s being dumped on us now,” Merchant said. “It’s an AI tool that helps us understand what that data actually means."