As AI agents begin to take on a greater role in fleet management, one message stood out from Mike Branch’s keynote at the 2025 Australasian Fleet Education & Leadership Summit: without clearly defined goals, even the most advanced AI won’t deliver meaningful results.
Branch, Vice President of Data & Analytics at Geotab, used his session to outline the rapid evolution of AI in fleet—from early observational systems to today’s intelligent agents capable of autonomous decision-making. But amid the excitement about predictive maintenance, driver coaching, and automated reporting, he warned fleet managers not to lose sight of the basics.
“In this world of agents, it will only try to achieve a goal,” Branch explained. “And so you have to have, as a fleet manager, clarity in what goals very specifically you’re trying to achieve.”
He illustrated this point with a common objective: improving utilisation. “You might say, ‘Okay, I’m trying to improve utilisation.’ Well, what is utilisation?” Branch asked. “Is it the number of active vehicles divided by the total number of vehicles? Okay. Well, what is active? Is active based on power on? Distance driven?”
The point is that a vague goal leads to flawed execution. If an AI agent doesn’t have a clearly defined target, it may draw the wrong conclusions or optimise for the wrong outcome. Worse still, it may generate recommendations that undermine the very efficiency the fleet is trying to improve.
This is particularly important in a modern fleet environment where AI systems aren’t just monitoring—they’re starting to take action. Geotab’s new ACE platform, for example, allows users to interact with their fleet data using natural language queries. Ask it about safety risks or maintenance needs, and it generates insights based on vast amounts of telematics data. But those insights are only as good as the definitions and context the system is working with.
“Very simple questions around improved utilisation as a goal can get complex quickly,” said Branch. “If you can’t measure every single piece, you’re going to have these agents go bananas.”
Setting clear goals also improves collaboration across departments. Whether it’s finance, operations, or maintenance, everyone needs to understand and agree on how key metrics are defined. If one team defines ‘downtime’ as time in the workshop and another as time off the road, any AI-based optimisation will be built on shaky ground.
Branch also connected this need for clarity to the broader issue of trust in AI. Agents are only useful if fleet managers trust them to deliver accurate and actionable insights. That trust begins with a clean strategy, well-articulated goals, and precise definitions.
“If we can do this right together… we really can unlock what’s next,” Branch concluded.
For fleet managers, the path forward is clear: before deploying AI tools or intelligent agents, take the time to set clear, measurable goals. It’s the foundation on which all successful AI strategies—and improved efficiency—will be built.
