All Ears: How This A.I. Startup is Helping Make Cars that Listen
Kim Lachance Shandrow | December 1, 2017
Imagine if your car had its own sense of hearing. It could warn you about engine and brake problems — before they become dangerous or too expensive.
If your car came equipped with artificially intelligent acoustics “listening” software, it could do just that. And it could automatically listen up for ambulance sirens, too — much better than you could — and tip you off to unsafe road conditions before you have to pump the brakes, or worse, crash into someone or something.
The idea of cars “listening” to their environment to make drivers, pedestrians and city streets safer is the core driving concept behind OtoSense, Inc. Sebastien Christian, a noted quantum physicist, and neuroscience and speech pathology expert, founded the Palo Alto, Calif.-based, 15-employee startup in 2014. Having closely worked with deaf children for a decade, he’s long been fascinated by the intricate ways in which listening helps us make sense of our world.
Christian’s fascination eventually led him to create an innovative artificial intelligence “machine hearing” platform that he claims can “make sense of out of any sound.” The special software can even learn and adapt to new sounds in complex environments on the fly, a skill that could potentially save driver and pedestrian lives.
We recently chatted with Christian, Otosense’s CEO, along with Thor Whalen, the startup’s chief technology officer. We wanted to find out how their futuristic “listening” tech can make driving safer, how it’s being used and what’s next for them. Here’s what we found out:
What was the moment like when the entrepreneurial lightbulb went off and you knew you had to launch OtoSense? Was there a defining moment?
Christian: “I was working on the brain side of the AI equation, and with kids with hearing loss. The idea came to me then, this desire to start a business that would have something meaningful to do with machine learning technology, that would help deaf children. Then we pivoted away from that moving away from children and exploring other avenues, leading us to where we are today.”
Can you take us through the basics of how OtoSense’s sound recognition platform works — and how can it make the world a safer place?
Whalen: “Our technology can learn to hear engine and other problems in cars in the way that humans do, but the technology listens constantly, and it can alert drivers to problems. This is somewhat like having a mini-technician in the car around-the-clock, hearing issues before the driver or anyone else even hears them. This is the great advantage of AI. It can listen constantly, 24/7.
This can also lead to time and cost savings, with technicians and mechanics no longer having to listen for car problems, for sometimes up to 45 minutes and longer. In other applications, like in factories, we are also saving on costs around machine maintenance.”
Christian: “As for what problems our technology solves, it’s a system that is able to learn, identify and name sounds — without any human ear telling it exactly what’s happening. It proactively listens for that weird sound that your car makes, or to the sounds that a plane or a train makes, to anything you can think of.
With our technology today, we can help plane-makers and automakers test vehicles better than before. We can help factory floor managers reduce the maintenance costs of their machines by identifying when the machine is about to go wrong.
We also have requests for our technology from healthcare companies, particularly with patient monitoring and detection around illnesses like COPD and sleep apnea. In these ways, we will transform the ways machine hearing can help people be not only safer, but healthier as well.”
How is OtoSense’s technology being used in the market now?
Christian: “We are working with Peugeot as a customer in France. LG and PSA Groupe are also partners using our technology. We also have customers in the industrial arena, but aren’t able to share their names at this time.”
What is the biggest challenge you’ve faced on this journey?
Christian: “The main challenge is that we are building something that is so complex and that has so many numbers of potential applications. And each time we talk about it, people say ‘Oh, you can do so many things! But will you succeed in doing just one?’ We’re taking efforts to prove that, yes, we can meet the needs of many applications, and can use our technology to power anything that can benefit from listening.”
What’s your advice for future entrepreneurs looking to launch tech startups?
Christian: “Bill Gates and Steve Jobs both said something like this, and it comes to me so often…It’s that you really have to love what you’re doing. You have to be so obstinate and perseverant. Any lack of passion can make you abandon your whole dream.”
Whalen: “My advice is don’t try to conquer the world immediately. Choose one battle at a time or nothing will get done.”
Whalen: “We will continue to push forward the idea that AI isn’t about imitating the human completely, but rather what is possible today with our technology, which is replicating human behavior, doing it 100 times in 24 hours, 7 days a week. I think we’re doing pretty well now, and that we will continue to develop and expand upon in the future.”