At the 2025 IPWEA Fleet Conference, one panel stood out for its grounded, real-world perspective on a topic that’s often wrapped in hype: artificial intelligence. Hosted by IPWEA Consultant Ken Goldberg, the discussion featured insights from across the fleet technology landscape. The panel included Darren Gore, General Manager at Summit Fleet Leasing and Management; Richard Biffin, Chief Technology Officer at Smartrak; Mark Hosking, CEO at Formbird Fleet; and Alkan Ciftci, Business Development Manager at Geotab.
Each speaker brought a practical mindset, focused on what fleet managers can do with AI now—not in five years’ time. Their message was clear: AI is not a revolution. It’s a tool. But used wisely, it can save time, reduce costs, improve safety, and help fleet teams make better decisions.
Getting Rid of Paper and Manual Processes
Darren Gore kicked things off by explaining how Summit Fleet has embedded AI into its operations to eliminate manual work. “We don’t want people stuck in spreadsheets,” he said. “Our focus has been on using technology to do the repetitive stuff so our team can focus on adding value.”
Summit uses AI to scan VINs against recall databases and send automated alerts to customers. They also validate odometer readings across multiple data sources—like fuel cards and telematics—and use AI to flag suspicious or inaccurate entries. These small wins add up to big savings in time and improve the quality of data fleet managers rely on every day.
From Reports to Conversations
For Richard Biffin at Smartrak, AI is about making data accessible to more people. Traditionally, fleets needed analysts or IT support to build custom reports. AI is changing that.
“We’re working on tools where someone can just type a question—like you would in ChatGPT—and get an answer,” he said. “It’s like having a Siri for dashboards.”
Instead of building 20 reports to cover every scenario, a generative AI model allows staff to ask questions as they come up: How many vehicles are due for replacement? What’s our average idle time? The goal is to support better decisions by making data more conversational and less technical.
Richard made it clear that these tools won’t replace human judgement. “AI won’t make the decision for you,” he said. “But it can give you the insights to make a better one.”
Better Driving and Smarter EV Choices
Geotab’s Alkan Ciftci shared real-world use cases that fleets are already benefiting from—particularly around safety and sustainability. “Driver risk is one of the first areas we’ve applied AI,” he said. “We’re analysing braking, acceleration and cornering to identify high-risk drivers and provide coaching feedback.”
Geotab is also using AI in its EV Suitability Assessment (EVSA) tool. It looks at trip patterns, range, dwell time and payloads to recommend which internal combustion vehicles can be replaced by electric or plug-in hybrid alternatives. “It’s not guesswork—it’s data-backed,” said Alkan.
For fleet managers under pressure to meet emissions targets, this kind of support tool can shortcut the EV transition planning process and help avoid expensive mistakes.
Use AI Like a Library Assistant
While many are excited about ChatGPT-style tools, Mark Hosking from Formbird Fleet issued a caution: know what the tool can and can’t do.
“Large language models are like very sophisticated parrots,” he said. “They don’t understand your fleet—they just predict what the next word should be based on massive data sets.”
Still, he sees real value in uploading your own content—like workshop manuals, internal policies or user guides—and using AI to help staff find information fast.
“One thing you could do today is load your manuals into a private AI model,” he said. “Then your team can ask things like ‘What does our tyre policy say about pressure checks?’ and get an answer instantly. It doesn’t replace human expertise—but it saves time.”
Data Governance Still Matters
With more fleets exploring AI, the panel warned that data security and privacy must be a top priority.
“You’ve got things like home addresses, regos, and driver behaviour in your datasets,” said Alkan. “If you’re using external tools, you need to know where that data is going.”
That’s why Geotab and Smartrak build AI tools inside their existing platforms—so data stays local, secure and under customer control. Microsoft and Google have also recently introduced controls to prevent user data from being used to train public models.
“Don’t just upload everything to a public tool,” added Richard. “Use platforms that offer data protection, or start with local models that run offline.”
Don’t Expect Miracles from Bad Data
A recurring theme throughout the panel was the importance of clean, structured data.
AI doesn’t work well with messy spreadsheets or incomplete records. “You still need good inputs,” said Darren. “If your fuel records are patchy or your vehicle list isn’t up to date, the answers you get back won’t be helpful.”
Richard recommended starting small—with a single dataset you trust, like emissions or service intervals. Then ask questions you already know the answer to. “It’s the best way to learn what the tool can actually do,” he said.
What Can Fleet Managers Do Next?
If AI feels overwhelming, the panel offered three practical steps for getting started:
- Identify a repetitive process – If there’s a task your team does every week that feels manual (checking fuel usage, scanning recall lists, formatting reports), it’s a candidate for automation.
- Upload your documents – Loading your fleet policies, safety guidelines, or equipment manuals into a secure AI tool can give staff an easy way to find answers and reduce training time.
- Experiment with your own data – Pick one dataset you trust and run it through a tool like Microsoft Copilot, Gemini or a private GPT model. Ask simple questions and test the results.
AI Is a Tool—You’re Still the Manager
As Ken Goldberg wrapped up the session, he reminded attendees not to wait for perfection. “AI is already here. You don’t need to solve everything overnight—but you can start now.”
Mark Hosking summed it up best: “We’re on day one of AI in fleet. It’s not magic, but it is useful. Just remember—it’s a tool. Use it where it helps, and don’t expect it to know more about your business than you do.”
Fleet managers may not need to be data scientists, but in 2025, they do need to understand what’s possible. Whether it’s reducing risk, saving time, or making the case for EVs—AI can be a valuable part of the toolbox.
