I once heard someone describe a veteran operations manager as “the second operating system of the business”.
At first, it sounded like a joke. Then I realised how accurate it was.
Every fleet has people like this. They know which routes usually get delayed, which drivers need more support, which charging locations are dependable, and which shortcuts are worth avoiding. When something unexpected happens, people instinctively turn to them because their judgement has been shaped by years of real-world experience.
That experience is valuable. But when too much of it sits in one person’s head, it also becomes a risk.
The issue is not that these people are too capable. The risk is that the organisation has become too dependent on their individual judgement, and that this judgement, however good, has never been captured in a form that can scale.
As fleets grow, the number of operational decisions multiplies quickly. Vehicle assignments. Route planning. Charging stops. Driver instructions. Operational priorities. What once relied on instinct and experience and instinct becomes increasingly harder to coordinate consistently across dozens or hundreds of vehicles.
Scaling EV fleets – simpler, or more complex?
For EV fleets, the challenge, can in some ways, become more complex.
Take charging, for example. Unlike conventional refuelling, charging decisions involve multiple trade-offs. Drivers and fleet managers must consider factors such as charger location, charger availability, charging speed, battery levels, varying charging costs, wait times and operational requirements.
Each of these variables shifts throughout the day, and the best choice depends on how they interact at a given moment. That is precisely the kind of multi-variable, real-time problem that a person can solve well a few times a day through experience — but not reliably, hundreds of times a day, across an entire fleet.
Multiply those decisions across an entire fleet and the challenge becomes one of consistency.
This challenge is not unique to fleet operations.
Airlines do not rely on experienced pilots to personally solve every situation. Manufacturers do not depend on factory veterans to manually maintain quality standards. Instead, they capture expertise and embed it into systems, processes and operating frameworks that can be applied consistently across the organisation.
Increasingly, those systems are powered by artificial intelligence — models that learn from vast operational histories and apply that learning uniformly, every time.
The same question now applies to fleet operations: how do you ensure good decisions are made consistently, even as fleets become more complex?
Making expertise scalable
The answer lies in digital systems and, increasingly, in the AI that sits within them.
Digitalisation is not about replacing experienced people. It is about transferring operational expertise from individual memory into systems that can support decisions at scale.
This is where AI changes what is possible. Much of a veteran manager’s value comes from pattern recognition — an instinct, built over years, for reading a situation and knowing what usually works. Modern AI is well suited to exactly this. Rather than asking managers and drivers to manually evaluate every variable, AI-enabled systems can analyse operational data, learn the patterns that lead to good outcomes, apply business rules and fleet priorities that managers define, and recommend actions based on real-world conditions.
The result is greater consistency across the fleet. Drivers are supported with better information, while fleet managers retain control over the priorities and rules that guide decision-making. The AI does not overrule human judgement; it extends it — turning the experience of a few into decisions that the whole fleet can benefit from. In this way, good judgement becomes embedded into the way the fleet operates every day.
Turning data into decisions
Consider these questions:
- Are fleet priorities such as cost, time, charger preference and vehicle readiness applied consistently across every journey?
- How much operational knowledge still exists only in the minds of a few experienced employees?
- If two drivers face the same low-battery situation, would they make the same charging decision?
What these questions reveal is that operational performance isn’t determined by a few major decisions. It is shaped by hundreds of small decisions made every day.
The challenge, therefore, isn’t building expertise. It’s scaling it.
As EV fleets become increasingly connected, they are generating more operational data than ever before. On its own, that data is just noise. AI is what turns it into foresight — anticipating charger availability and wait times, flagging vehicles likely to fall short on range, and recommending the charging or routing decision that best fits the fleet’s priorities before a problem arises. At the same time, advances in analytics and AI are creating more opportunities than ever to turn that data into practical operational insights.
The organisations that succeed will be those that combine these AI capabilities with the experience and judgement accumulated by their fleet managers and operators. Together, they can identify what works, capture that knowledge, and apply it consistently across the operation. The human sets the priorities and knows the context; the AI applies them tirelessly, at a scale and speed no individual could match.
Digitalisation, and the AI within it, gives fleet operators a way to preserve that knowledge, structure it, and apply it more consistently. Not as a replacement for experienced people, but as a way to make their expertise work harder across the entire fleet.
That is what separates fleets that depend on heroes from fleets that build capability.