For many companies, saying fleet safety is important is a lot like saying people are our most important asset. It sounds good, but it seldom matches reality. That may be changing. For a growing number of forward-thinking companies, fleet safety has become about more than simply responding to accidents after they occur or focusing on compliance. They are using data, technology, and AI to identify potential risks, improve driver performance, and prevent incidents before they happen.
What separates these companies from the rest, and what lessons can we learn from them? That’s the topic I discussed with Hayden Cardiff, Vice President of Safety Solutions at Descartes, during a recent episode of Talking Logistics.
A Safety Culture
I began by asking Hayden what differentiates companies with a strong safety culture from those that simply meet minimum safety requirements. Hayden says companies with a strong safety culture embody safety from the top down and the bottom up. “That takes a lot of energy and effort. It takes the ability to understand that it’s everyone’s job, and it can’t be just lip service.”
In practice, he says, you can’t have a safety manager telling a driver to slow down while an operations manager is congratulating that same driver for getting freight delivered on time — even though the driver may have had to speed to make the delivery. “You have to have everyone working hand in glove… and have that same message pervasive throughout the organization.”
Hayden notes that what separates the leaders is going beyond standard industry practices. As an example, he says many fleets invest in cameras and then coach drivers after individual safety events occur. But over time, that can become white noise to drivers or even frustrate them if they feel they’re constantly being nagged. Instead, leading companies use technology to tie together all available data to better understand driver risk and become proactive and predictive in reducing risky behaviors.
Turning Insights into Action
How can companies use technology to turn safety data into actionable insights? Hayden comments that the trucking industry has a phenomenal amount of safety-related data available. “We have a lot of data but not a lot of insight.”
Hayden says cameras and video telematics are now table stakes. Companies should also incorporate ELD data, FMCSA inspection and violation data, as well as internal information such as customer feedback, learning management systems, coaching records, and other safety and operational performance indicators.
He adds that the key is having technology that can sort through all of this data, identify patterns, and distinguish signal from noise. This is where machine learning, AI, and predictive analytics come into play. By adding context to the data, they help identify the leading indicators that precede crashes and enable companies to move from reacting to incidents to proactively managing risk.
Predicting and Preventing
How do companies move beyond reacting to data to predicting and preventing safety issues?
Hayden notes that this is where AI and machine learning provide tremendous value. This technology analyzes driver behaviors and characteristics that have led to crashes in the past and uses those patterns to predict statistically which other drivers may be at greater risk in the future, enabling companies to intervene proactively. Importantly, Hayden notes that the analysis is not limited to a company’s own drivers. The technology analyzes anonymized data across many fleets to create a much broader view of driver behavior and outcomes.
Hayden explains why this is so important: “Crashes are a very infrequent occurrence… Being able to predict crashes is challenging, if not impossible, with the level of data that any one individual fleet has.”
He adds that Descartes has more than 40 billion miles of driver data and more than 500,000 preventable crashes in its dataset. “If you don’t have a large amount of robust data around crashes, you can’t predict crashes.”
The Role of AI
No discussion of technology today is complete without addressing the role of AI. Hayden made an observation I found particularly interesting: “AI… has to be there to help enhance human-to-human engagement.”
He explains that too often we think of AI replacing people. While AI can certainly automate routine tasks, in the case of driver safety it can identify the small percentage of drivers who present the greatest risk and provide safety and operations personnel with the context and coaching recommendations they need to have more meaningful conversations with those drivers.
Another example is using AI-enabled in-cab cameras to detect when a driver picks up a phone and immediately prompt them to put it down. Ultimately, Hayden says, AI should help organizations spend less time managing data and more time helping drivers improve their performance.
Looking Forward
With technology changing and advancing so rapidly, how can companies position themselves today — and in the future — to improve their safety programs and outcomes?
Hayden shared many valuable suggestions centered on the idea of “future-proofing” your organization, as well as many additional insights I didn’t have space to include here. I encourage you to watch the full episode to hear his comments and expertise firsthand. Then keep the conversation going by sharing your own thoughts and questions on this topic.
