Proactive fleet strategies: Leveraging AI, integration, predictive maintenance for future success

CLEVELAND—As fleet leaders and the infrastructure teams that support them look toward the next decade, the traditional "firefighter" mentality—where managers spend their days reacting to the crisis of the moment—is being replaced by a sophisticated, proactive approach rooted in connected operations. The concept isn't new, but holistic integration of data into forward-looking operations will be essential in the very near future.

That’s the takeaway from a panel discussion, “Is Your Fleet Ready for 2030?” featured at this year’s NAFA Institute & Expo here. Moderated by Dan Franck, head of global partnerships at Samsara, industry leaders from the public and private sectors shared a clear consensus: The future of fleet management lies in transforming vehicles into "data platforms" where real-time insights drive immediate action.

The catalyst: AI camera systems, behavioral change

Fleet safety is the essential measure for operations, and connectivity is the key from here on out. The "aha moment" for Paul Palmer, senior director of safety for national tree care company SavATree, came when his fleet was piloting AI-based camera systems. Initially skeptical of the data burden, Palmer discovered that in-cab coaching features allowed the technology to manage driver behavior automatically—that is, without intervention from local managers. Basics such as seatbelt violations, harsh braking, and distracted driving dropped significantly as drivers received real-time "nudges.”

“Having the data and not acting on it was a huge risk, in my opinion,” Palmer said. “The technology can manage this on the front end. The old state of affairs, before you had in-cab coaching, a driver could have a violation on a Tuesday and it'd be Wednesday morning before their manager or supervisor is pulling them into the office and saying, ‘hey do you remember this yesterday morning?’ Well, no, the driver didn't remember it then, so they're immediately defensive.”

Jon Malish, public works manager for the City of Akron, echoed this success.

By utilizing AI to track egregious speeding—running 20 mph over the posted limits—his department saw violations drop from 393 events in a 90-day period to zero after turning on in-cab alerts.

“The drivers felt their privacy was being violated; Big Brother was watching them. And what were the effects of that of that violation going to be?” Malish explained. “Fixing the behavior, that's all that we're trying to do with AI. And it's been absolutely fantastic seeing our cases drop.”

Beyond behavior modification, these systems provide a vital shield against liability. Malish cited a case where forward-facing footage from a street sweeper exonerated a driver by proving a vehicle’s mirror was already damaged before the crew arrived.

This "exoneration factor" is a game-changer for fleets with "deep pockets" and “big-name logos” that are often targets for litigation.

Beyond the dashboard: the power of integration

While cameras are a potent tool, the panelists emphasized that the true value is unlocked when the data is integrated into a broader fleet management system. 

Steve Jastrow, Element Fleet Management SVP, advisory and analytics, pointed out that having a "pretty dashboard" is not enough to satisfy insurers or drive ROI. To see results, data must be connected to the entire ecosystem—from maintenance and fuel to payroll and risk management.

“There's no real standard way of measuring safety statistics across the industry,” Jastrow said. “Don't get so focused on making sure you have the right piece of data coming that you ignore the other data elements. Fleets implement, say, telematics but then they don't coach their drivers, they don't tie it to incentives or consequences so they're not integrating it into their operation. They'll never see ROI, no matter how many data points are coming in, if they're not actually using the data.

“Make sure you've got the right metrics in place and that they're connected through the ecosystem, so they're getting back to those who are actually operating the vehicle—that’s the carrot and stick—that's where you're going to ROI. Historically, we would just manage the vehicle and not the driver, but the driver is far more impactful on your cost performance than the vehicle itself. Telematics gives you the ability to manage the driver, not just the vehicle. Rather than just waiting for that perfect data to come in and then put it on a pretty chart and pin it on a wall—and hope everybody's going to understand what to do—that probably will not work.”

SavATree’s results support this integrated approach. By using APIs to link telematics data with their fleet information system, geo-fencing, and payroll, they were able to measure "dispatch and return" times that were previously invisible.

Managing just six minutes of wasted time per day across their thousands of drivers adds up to nearly $1 million in annual savings.

Similarly, the City of Akron integrates its telematics with Esri’s GIS platform to provide public-facing transparency, allowing residents to see exactly when their streets were plowed or their trash was collected.

Shifting from reactive to predictive

This connectivity is also revolutionizing vehicle lifecycles.

Jastrow highlighted the shift from "preventative" to "predictive" maintenance. Instead of relying on lagging data or simple averages, fleets can now use real-time engine diagnostics and odometer readings to better forecast when a vehicle needs service or replacement.

“The best way to avoid downtime events is to make sure that your vehicles are properly maintained,” he said. "It also gives visibility to when the vehicle needs to be replaced.”

At SavATree, this data-driven foresight allows them to rebuild sprayer pumps at the one-year mark because they know, statistically, that failure rates spike in year two. While this requires upfront capital, it eliminates the "killer" financial impact of downtime during their busiest seasons.

Furthermore, utilization data is allowing managers to "right-size" their fleets; rather than purchasing new assets for a new contract, they can identify and transfer underutilized vehicles from other regions.

“You're going to where the puck is going rather than constantly chasing it,” Jastrow said. “So it's really shifting how we think about managing fleets in today's world

The road to 2030

As fleets face increasing pressure to adopt electric vehicles and meet sustainability targets, the panelists advised a "surgical" rather than a blanket approach.

Rushing into electrification without analyzing route data is a path to operational disappointment, Jastrow cautioned, noting that extreme weather and unpredictable long-distance trips can "wreck a business" if not accounted for.

For heavy-duty operations where electrification may not yet be viable, Malish suggested using telematics to enforce anti-idling policies as a more immediate way to lead on sustainability.

The closing advice for fleet managers feeling overwhelmed by "data paralysis" was simple: start now.

"Don't wait until the data is perfect," Palmer urged, noting that waiting only puts a fleet further behind the curve.

Whether the goal is safety, cost reduction, or better utilization, the key is to pick one "North Star" metric—such as cost per risk mile—and build the data ecosystem around it. By 2030, these connected systems won't be optional add-ons; they will be the standard for efficient—and cost-effective—fleet operations.