K&B Transportation Uses AI Cameras, ChatGPT for Driver Coaching
Mid-sized carrier deploys AI-powered safety cameras and generative AI tools to review driver behavior, but keeps humans in the loop to catch data errors and verify coaching accuracy.
How is K&B Transportation using AI in its safety program?
K&B Transportation is running AI-powered safety cameras in its trucks and using ChatGPT to help generate driver coaching materials, according to Lance Evans, a company representative who discussed the fleet's AI deployment in a recent industry podcast. The carrier uses AI to analyze camera footage for safety events and to draft coaching content, but maintains human review of all AI outputs before they reach drivers.
The approach reflects a broader shift in how mid-sized fleets are adopting AI tools. The technology handles initial data processing and content generation, while fleet staff verify accuracy and context before acting on the results.
What AI tools is the fleet running?
K&B operates AI-enabled safety cameras that flag driver behavior events automatically. The system reviews footage and identifies incidents that may require coaching or follow-up. Evans noted the fleet also uses generative AI tools like ChatGPT to help produce coaching materials and other operational content.
The camera systems represent the hardware side of AI adoption in trucking. These units mount in the cab and use onboard processors to analyze video in real time, tagging events like hard braking, lane departure, or following distance violations. The AI categorizes the event and sends flagged footage to fleet managers for review.
Why does K&B still require human oversight of AI outputs?
Evans emphasized that AI-generated data and content require human verification. Fleet staff review camera footage flagged by AI to confirm the system correctly identified a safety event. They also check AI-generated coaching materials for accuracy and relevance before sending them to drivers.
The reason is straightforward: AI systems make errors. A camera may flag a hard-braking event that was actually an emergency avoidance maneuver. Generative AI may produce coaching language that is technically correct but misses context specific to the driver or the incident. Human review catches those gaps.
This mirrors what other fleets have found when deploying AI tools. The technology reduces the time required to process large volumes of data, but it does not eliminate the need for experienced fleet managers to interpret results and make final decisions. K&B Transportation blocks mobile phones and geofences speed zones as part of a broader safety technology strategy that combines hardware controls with AI-assisted coaching.
What role does AI play in driver coaching?
K&B uses AI to help structure coaching conversations based on camera footage and other data. The system can draft initial coaching points or summarize patterns in a driver's behavior over time. Fleet managers then review the AI-generated material, adjust it for accuracy and tone, and deliver the coaching.
The workflow speeds up the coaching process. Instead of starting from scratch with each incident, managers begin with an AI draft that pulls relevant data points. They verify the facts, add context the AI missed, and personalize the message. The result is faster turnaround on coaching without sacrificing quality.
Evans noted that fleets need to embrace AI technology to stay competitive, but the human element remains critical. The technology handles repetitive data analysis and content generation. People handle judgment calls and relationship management with drivers.
What does this mean for small fleets considering AI tools?
K&B's approach offers a template for how smaller carriers can adopt AI without overhauling their entire operation. Start with specific use cases where AI handles high-volume, low-complexity tasks (reviewing camera footage, drafting routine communications). Keep experienced staff in the verification and decision-making loop. Treat AI outputs as drafts, not final products.
The hardware cost for AI-enabled cameras has dropped enough that fleets under 50 trucks can justify the investment if the system reduces coaching time and catches safety events earlier. The key is setting up a review process that catches AI errors before they reach drivers or affect CSA scores. A camera that flags false positives wastes manager time. A coaching message with factual errors damages driver trust.
Fleets already running standard dashcams should evaluate whether their current vendor offers an AI upgrade path or whether switching to an AI-capable system makes sense at the next contract renewal. The ROI depends on how much time managers currently spend reviewing footage manually and how many safety events go uncoached because staff lack bandwidth to process all the video.
For generative AI tools like ChatGPT, the barrier to entry is lower. Fleets can test these tools for drafting safety bulletins, summarizing incident reports, or generating training outlines without hardware investment. The same rule applies: verify every output before it goes into production use. The technology is useful when it saves time on tasks that do not require deep expertise. It becomes a liability when fleets skip the verification step and assume AI-generated content is always correct.
