Brian Hearing is an engineer and acoustics expert. He previously worked for the Department of Defense on sound-related projects, such as technology that can listen for distant tanks or helicopters.
Now Hearing is exploring the civilian aspects of sound. In particular, he hopes to create waves with BrakeAudit, a company he cofounded.
The company aims to take advantage of a basic fact about commercial vehicles: They’re loud. All that noise contains a wealth of information. By analyzing the sounds, BrakeAudit can identify any issues with the brakes.
“A law enforcement officer knows reasonably well which trucks he’s going to find problems on by listening to them, sometimes by looking at them. A mechanic, the first thing he asks you, ‘What does it sound like?’ What we’re doing is automating that process and embedding the knowledge human experts have into a computer so that the entire system is fully automated,” Hearing says.
The company is still in its early stages. Two patents have been applied for and the equipment is up at partners’ facilities so that BrakeAudit can fine-tune the system.
The following conversation has been lightly edited for clarity.
TBR: How does it work?
BH: It’s relatively straightforward. The good thing about acoustics is that they’re really affordable compared to other technologies. The processing capabilities in a cell phone are orders of magnitude more than we need.
So what we do is we put up affordable sensors at a location where vehicles are coming to a stop. Some locations are exits of a bus depot or a maintenance yard, a warehouse where trucks are coming in and out, coming to a complete stop before going onto the main road. We set up two sensors, one on each side. The system listens to the sound of the vehicle, runs the algorithm, and reports it back to maintenance
We have multiple different acoustic signatures of problems that we look for.
In a maintenance yard where you’re seeing the same vehicle over and over, we can compare it to what we saw last time. If we haven’t seen it before, there are other signatures we check for to see if they match the sound the vehicle is making.
There’s other things we’re finding too. Brake line leaks are something we can detect, which was a surprise to us. Further down the line you can apply it to the entire vehicle. The engine makes certain sounds, the exhaust system. But we’re really focused on brakes at this point.
TBR: How do you communicate the results?
BH: Different users would receive different information.
Law enforcement would receive a text, basically saying, ‘Hey, pay attention to this one.’
For a maintenance manager, we send out a variety of different information. If it’s something we find that’s really bad, we’ll send out an alert, saying, ‘Have that truck or bus pulled immediately. It shouldn’t be on the road.’
For less urgent problems, we report back to the maintenance manager. That can be done both with a periodic email or an online system they can log into.
TBR: You must have required a big database of sounds to start identifying these signatures, right? How did you amass that database?
BH: We’ve partnered with different fleets and law enforcement organizations to collect data.
We’re still in a development phase. So we’re collecting as much data as possible and then verifying what we’re seeing in the field with the maintenance people. This is very much a cooperative event at this point.
We attend law-enforcement events, when a truck gets pulled over at a weigh station and they do the inspection, so we’re also collecting those results as well.
We’re also talking with brake manufacturers. They run new designs in a controlled environment with a nice dynamometer, and they can run their components through these different tests, and that basically gives us the information ahead of time, before we see it in the wild.
TBR: How much of the sound that you analyze is audible to the human ear and how much is not?
BH: We’re finding that there’s a lot of content in the ultrasonic realm. A lot of vehicles put out audio emissions above what humans can hear. These can give you a warning that something is coming. We can usually identify problems before a human can hear it.
TBR: How close are you to a wide launch of BrakeAudit?
BH: The current plan is that probably in six months the product will be reliable enough to where it’s working on a fully automated basis to detect a handful of significant problems.
Right now the market that seems most interested is law enforcement. They’re looking at subset of problems compared to the maintenance manager.
Law enforcement is looking for things that could place a truck out of service, things like adjustors that are out, serious problems with the brake, wheel-end contamination, loose components.
Focusing on that subset of problems that law enforcement is interested in is easiest for us in the short-term.
TBR: What are the biggest hurdles to clear?
BH: Handling the wide variety of different types of brakes that are out there now is a challenge. Choosing what to focus on first and the different brake problems is critical.
It would be easy for us to bite off too much and say we can do everything for every vehicle. That’s just simply not possible at this point.
There are things that we feel like we can explore but we haven’t yet.
If regenerative brakes really take off, there’s going to be much less friction-based braking, so that could detract from the information we’re able to provide the maintenance manager. At the same time, autonomous vehicles are really interesting for us. If you take the human driver out of the loop, the maintenance manager has less information, so could be more reliant on technology like ours.
TBR: What’s your ideal scenario for growth for BrakeAudit?
BH: The market that is showing the most interest right now is law enforcement. We feel the next market after that will be fleet managers. They will want to know why they’re being pulled over, what is this new magic technology that officers are using to flag vehicles? That’s going to be the fastest expansion for us.
After that, I can see moving into passenger vehicles. At that point, there’s potential for using our technology onboard the vehicle, listening to the sound of the brakes from inside the vehicle instead of outside. So moving on to OEM onboard equipment would be the ultimate end goal.