The rise of artificial intelligence tools has been so rapid, and the potential uses are so widespread, the idea of adopting AI at your trucking company can seem intimidating or overwhelming. But it doesn't have to be, said panelists in a session during the Wex OTR Summit in early October.

This is a journey, a marathon, and we're still at the beginning of it, explained Robert Nathan, founder and CEO of Envoy AI, which builds autonomous agents for logistics service providers.

Nathan advised fleets to start small, identifying time-consuming processes to find tasks that AI can improve or repetitive tasks you can take off a human's plate.

“Take an area where you’re seeing a lot of people spending time, and let’s start to see what they’re doing. Let’s unpack that," he said.

What are the workflows involved? And is there something in that workflow that AI could do?

The panel emphasized that just using artificial intelligence just to use it is not the right way to go about it. “AI-powered” is not necessarily a selling point for truck drivers or other workers. They want to know how it can make their lives better.

Getting Truck Drivers On Board with AI

J.P. Gooderham, CEO of Storyboard, whose company offers a hands-free voice assistant for truck drivers, emphasized the importance of getting driver buy-in on artificial intelligence by focusing on what it means for them.

And that doesn't mean just showing them how cool it is that they can talk to an AI agent while they're driving to find a fuel stop that takes their fuel card, near a restaurant serving their favorite foods, while optimizing their hours of service and fuel savings. (Although it doesn't hurt.)

“Going in and telling your drivers, ‘We have an AI product,’ is not a sale. It's not showing them what they're going to get, what's going to be the opportunity for them.”

He said fleets should involve drivers early — through surveys and advisory councils — and frame AI tools around solving their real problems.

Deploying AI Effectively: Training How to Use Generative AI Prompts

Jake Fields, co-founder of Platform Science, said companies should have a program in place to make sure that people are not only using AI, but also that they are using it correctly.

He advised fleets to train employees on prompt-writing, since better questions yield better AI results.

“The AI is only going to give you the best answer for the question you ask,” he said.

He suggested using AI to help create better prompts, which may seem counterintuitive, yet can be very effective.

“Before you say, ‘How do I solve this problem,’ ask the AI, ‘How should I put together a prompt for solving this problem?’” Then use or refine that prompt to ask your question and you'll get a much better answer, he said.

“That one little step we've seen actually really improve the performance and utilization of AI internally across different sectors of the company.”

Chris Oliver, chief marketing officer of Trucker Path, said there's been a push at his company to adopt AI, even if it’s something as seemingly simple as using ChatGPT or another generative AI service to help make decisions. He uses it in this way frequently.

But he also cautioned about the accuracy of these tools. 

“Let's be honest, it's a lot of fancy Google searches that generate the data and responses that you get with AI, and it's a lot of data, but it's not always contextually accurate," he said, saying humans still need to review the results.

Fields added that overly broad prompts often lead to “hallucinations,” when a generative AI model creates false or nonsensical information and presents it as fact. So learning to create precise prompts is key.

Protect Your Trucking Company's Data When Working with AI Providers

Panelists cautioned about the importance of protecting your data when adopting AI tools or choosing AI technology partners to work with. 

“There are a lot of AI … companies that are asking for carrier and broker data,” said Nathan. “Protecting that data is critical. 

“A lot of these AI companies are raising a lot of money by saying they're going to disrupt trucking and the brokerage industry,” he said. He warned fleets to vet AI vendors carefully, asking why they need certain data and how it will be used.

Fields added that fleets should share only the minimum data required for a given task.

“Does a given application need to have driver information, or do they just need to know point A to point B?”

Gooderham noted that strong data protection also builds driver trust. “We can look drivers in the eye and say, ‘We are being careful and we're protecting your data.’”

Focus on Small Wins

At Trucker Path, departments share weekly examples of how they use AI tools — even small ones — to encourage adoption, said Oliver. 

Gooderham said small “wins,” such as using AI to create better slide decks for presentations, can build confidence and momentum for adoption.

In fact, that example was one of the first ways he used AI. “I’m terrible at wording,” he admitted.

“If you’re trying to drive more AI adoption within your organization, share the wins.”