Travelers harnesses tech to lower trucking, manufacturing risks

di-travelers-031317
The Travelers Insurance Co. office building in Hartford, Connecticut in 2015.
Ron Antonelli/Bloomberg

Digital Insurance spoke with Chris Hayes, assistant vice president of risk control for workers compensation and transportation at Travelers Insurance. At first glance, it may not seem like workers' compensation and transportation have a lot in common, but workers' comp can be a big concern for insuring truckers and transportation fleets. With one of the largest risk control organizations, Travelers has devoted significant resources to trucking, workplace ergonomics, machine learning and managing returns of injured workers.

How do trucking and workers compensation go hand in hand?

Chris Hayes - Travelers.jpg
Chris Hayes, assistant vice president of risk control at Travelers.
Jeff Yardis
Both deal with safety management. They deal with people. They deal with behavioral safety. They have a lot of the same consulting methods. Ways to understand risk work for both disciplines. Motor vehicle crashes are one of the most significant sources of severe workers compensation losses. Having the two aligned really does work for us. Having them under one coordinator just makes sense.

What's the promise of telematics to reduce risk in trucking?

The promise is great. The reality is more challenging. Telematics gives trucking companies an opportunity to dig into individual data and behavior-based safety, really get into the individual behaviors of drivers on the road in the way we might in a factory or warehouse where you can observe workers on a regular basis and directly coach them.

Now it's much easier. The information's right in your hands. The challenge is that it can often be too much information. There's so much information that it is hard to go through it, sort through it, know what's important, know what's not important, and then be able to effectively manage the coaching for it. While it sounds great in theory, the application of it still has a lot of potential, but we find it often takes more effort than companies realize to really put it all together.

How did telematics for trucking become the source of so many data points?

Data quality has improved significantly. The ability for the data to be transmitted has improved significantly. It's not only the data coming out of engines and making that connectivity, but improvements in accelerometers to more accurately understand and predict vehicle motion. 

The advent of the reactive camera that can take an accelerometer trigger to save video, then moving machine learning into those cameras makes them even more predictive about what to keep and what to monitor.

The data just keeps on increasing, increasing, increasing. The challenge then is, there's so much, what do I do with it all? Making the information more usable and digestible for a fleet manager to change behaviors is critical. When talking to insurers about telematics, we spend less time talking about the specific technology. We spend more time talking about what's in the dashboard, what information they're getting, how they're using it and whether they have the right internal structures for safety management and communication.

Adding new features to the telematics is great and certainly will keep coming. We're focused on getting it down to just the usable information that will drive behavioral changes. If I have 500 vehicles on the road, I need to be able to see all 500 at once and see my top performing of the 500, the bottom performing of the 500, what specific behaviors do I need to coach for that bottom, what specific behaviors do we need to reward for the top, and how do I roll that together and communicate that to every driver.

Are these technologies used in telematics, such as machine learning and video, being applied in other environments?

Absolutely. AI-enhanced video is being used to look at tasks in manufacturing environments, construction environments and different industrial environments, to understand the ergonomics behind repetitive motion and lifting, and get very precise information on how employees might be injured because of these tasks. It's taking lessons learned about industrial safety and applying a different lens to make it more understandable and usable. 

We're spending a lot of time on retention of employees in the workforce and where the next generation of workers is going to come from. We have an idea we call "corridor of care." As employees are injured, there is a process from the first time you hear about the injury through to transitioning them back, first on transitional duty and then back to full employment. For every employer that says they can't find enough employees, it's worth asking how they are supporting the current workforce. We call this our workforce advantage process, engaging the current workforce and using the insights to build a better training and onboarding system.

In trucking, what challenges does automation create for working with risks?

There's a range of opinions about the long range vision for full automation. Odds are pretty good that we all will drive something with adaptive cruise control, automatic emergency braking and things like that. The next level is when I could get into my car, take a nap, and wake up at work. I don't think that will happen in my commuting future. 

If you can reduce crossing lanes unintentionally and reduce the risk of rear-end crashes, it can improve all of our lives. Smaller incremental developments of technology on the way to full automation are the really exciting things happening in the next few years. 

In the trucking space, it's more challenging because of the braking dynamics for trucks. Cars typically use hydraulic brakes. Trucks use air brakes. Air brakes take longer to engage. To rely on an automatic air brake is more challenging and will take more time to develop. But if you can slow down the vehicle ahead of a collision, to reduce the amount of force transferred in a collision, that is a pretty good brake improvement.

How can automation reduce risk in manufacturing or workers compensation?

As we go to more automation, robots and cobots [collaborative robots] and things like that, the way that the individual worker in a manufacturing environment engages with the work is going to be different. The skill set is different. The level of understanding of technology and just simple math is now different. The other real challenge as you go into a facility that is highly automated, you might not be running a machine, but monitoring a machine that is automated and engaging at certain points. How to keep that employee engaged, aware and focused on quality and safety when so much is automated is one of the challenges coming up for manufacturing.