What Chefs Want reduced accidents by 19% with AI Dash Cams and Automated Coaching

Building a proactive safety program that brings drivers into the conversation

After evaluating several providers, WCW chose Samsara. "Simply put, we went with Samsara because it's the industry-leading standard," said Eric Hailey, Regional Safety Manager. The initial rollout equipped the 600-truck fleet with dual-facing AI Dash Cams, giving the safety team its first real view of what was happening on the road. Real-time In-Cab Nudges give drivers that same awareness, flagging risky behavior in the moment so they can self-correct before it reaches a manager. But getting cameras into 600 trucks across 36 locations took more than installation. Safety leaders walked drivers through the dashboard itself, showing them no one was watching a live feed and that audio was off—answering privacy concerns before they became objections.

Coaching also followed a standardized format across all sites: each session touches on what the driver did well, what needs work, and what to focus on next. With AI-guided coaching sessions, managers are provided with talk tracks, recognition opportunities, videos, and more to help structure better training conversations. "Having the data to support coaching conversations is really effective," said Hailey. "We have real, raw numbers at our fingertips, and the data doesn't lie." Because that workflow lives in Samsara, the structure holds regardless of who runs the session. "Now we know every driver is getting coached properly and treated fairly—the process in Samsara is the same every time," said Hailey.

Building on that foundation, the team rolled out Samsara Coach Recaps—AI avatar coaches—for engaging, video-based feedback in the Driver App. Using Samsara's preset virtual coaches, each session walks a driver through the specific moments that affected their score and where to focus in the week ahead—individual coaching for every driver, at a scale three people could never cover on their own. Today, WCW runs Competitions, Recognition, Automated Coaching, and Agent Studio on the same platform.

A complete safety ecosystem drives a 19% accident rate reduction

Organization-wide changes showed up fast. Over a six-month period, the fleet saw an 86% decrease in heavy speeding, a 94% decrease in rolling stops, a 64% decrease in seatbelt violations, and more—contributing to a 69% overall drop in safety violations. Accidents fell 19% across five locations, and claims for serious crashes declined sharply. Video evidence also regularly proves driver innocence, protecting drivers from false claims. "We've had drivers who were doing everything right, and another driver still caused the accident. Having the footage to bring to insurance and challenge those claims has been huge—we're at 30 not-at-fault accidents this year," said Hailey.

That visibility also reshapes day-to-day culture. Sites post safety scores publicly, and "Safety Score Showdown" Competitions—driver contests set up by admins in the Driver App—have turned around safety and performance by channeling friendly, peer-to-peer rivalry using live leaderboards and rewards. Recognition includes Kudos sent via the Driver App, milestone celebrations, swag, and career opportunities for top drivers. "Samsara data doesn't just power our recognition program—it's how we identify our best drivers and promote them into trainer roles. It shows every driver there's a real path forward," said Hailey. Those trainers now lead onboarding for new hires, setting the standard from a driver's first weeks on the job.

Behind the scenes, that data also powers the administrative side of the program. A custom AI agent in Agent Studio now generates executive-ready safety summaries without manual work. "I used to plug data into spreadsheets and create graphs and charts. With Agent Studio, I don't have to do any of that. It easily saves six hours each week, and I don't need a PhD in Excel," said Hailey.

Today, safety at WCW runs on one system—and the team is already looking at extending what it can do with the platform next.