Art Credit: Greer Freshwater Burton
Scene: Mother Nature has pounded the Upper Midwest with snowstorm after snowstorm, blanketing the streets and causing massive traffic jams. At the end of every blizzard, snowplows and workers spreading a de-icing mixture of salt, gravel and sand appear like clockwork to clear the roads. All too often, though, ice reappears after frigid nights.
On a four-lane highway outside Minneapolis, cars going 55 mph slip and slide over a hidden ice patch. Vehicles drift dangerously into adjacent lanes. Twenty-five minutes later, a car swerves into oncoming traffic, badly injuring four and closing the highway for hours. Authorities report similar accidents in Cleveland, Milwaukee and Detroit.
A massive one suddenly opens up on a busy stretch of Olympic Boulevard. Drivers passing over it feel jostled and unnerved. A few automobiles are thrown out of alignment. An hour later, a speeding Porsche runs over the huge crater, jumping the sidewalk and slamming into a tree. The driver lives but suffers serious internal injuries. Traffic comes to a standstill.
The above scenarios are all too common. However, USC Viterbi faculty member Bhaskar Krishnamachari hopes to make them a thing of the past. The 35-year-old associate professor of electrical engineering has spent nearly a decade working on algorithms and software to make it possible for cars to “talk” to one another by sending messages through an ad hoc wireless network.
“I see this as a kind of early warning,” said Krishnamachari, who in 2011 the MIT Technology Review named as one of the world’s 35 top innovators under the age of 35 for his work on next-generation wireless networks. “If I know what’s coming up ahead of me, I can be prepared for that and slow down, take evasive action and, if I know about the problem a few intersections ahead, even take a different route.”
As envisioned by Krishnamachari, sensors in vehicles would alert internal computers of potential trouble. The computers would then process the information about potholes, icy roads and the like, which would be shared with nearby vehicles over radio waves.
Bhaskar Krishnamachari
To prevent drivers from being inundated with messages, Krishnamachari has designed algorithms that would aggregate warnings to limit the number commuters receive. Other algorithms could determine how long and where such alerts would be disseminated.
And temporary peer-to-peer networks, Krishnamachari added, won’t burden the already stretched existing cellular networks.
In addition to their safety implications, Krishnamachari’s temporary network of clustered cars has entertainment value – literally. With vehicle-to-vehicle communication, a car could use the network to download a movie or videogame for passengers and then share it with nearby automobiles. Interestingly, a vehicle could theoretically learn its users’ tastes and preferences and automatically download relevant content based on that profile, he said.
Krishnamachari began his talking car research in 2004 in a collaboration with Ghandeharizadeh. Krishnamachari later partnered with Fan Bai, a researcher at General Motors who earned his Ph.D. at USC Viterbi's Ming Hsieh Department of Electrical Engineering. GM has supported the pair’s research since 2008. Bai and Krishnamachari also jointly received a National Science Foundation grant titled “Information Centric Networking on Wheels (ICNoW) – Architecture and Protocols.”
“I believe Bhaskar’s academic work is pioneering and contributes much to the research community in the vehicle-to-vehicle networking field,” Bai said.
The U.S. Department of Transportation and several car manufacturers are currently conducting similar research on vehicle-to-vehicle communication.
So when might Peugeot say “Bonjour” or an Audi utter "Guten Tag?” Krishnamachari believes the technology may not appear in cars on a large scale for at least another decade.
There is the technical challenge of security. Hackers might compromise peer-to-peer networks and spread false and possibly even dangerous information to drivers. This is something that car companies and academic researchers are still working to address. Additionally, there is a business challenge for car companies in figuring out whether they want to become first-movers in the space. (Given the safety benefits of vehicle-to-vehicle technology, the government might eventually mandate its installation, experts said.)
Krishnamachari thinks fleet vehicles, such as groups of cars or buses owned by businesses or the government, are likely to lead the way “because it’s easier to manage deployments and upgrades.” For a vehicle-to-vehicle system to operate efficiently and successfully, he believes about 10,000 cars in Los Angeles must be equipped with the technology.
Krishnamachari grew up in India. He moved to New York City at the age of 15, when his auditor father landed a position at the United Nations. He earned a B.E. in electrical engineering from The Cooper Union for the Advancement of Science and Art in New York City, and an M.S. and Ph.D. from Cornell University.
Coming to USC in 2002, he won a prestigious NSF Early Career Award in 2004 for his work in wireless sensor networks. In 2005 Krishnamachari was appointed the first holder of the Philip and Cayley MacDonald Early Career Chair at USC, and three years later, he became the inaugural Ming Hsieh Faculty Fellow in Electrical Engineering.
Most recently, Krishnamachari has been working on using swarms of robots to provide wireless connectivity in remote, infrastructure-less environments. So innovative is his research that he was recently invited to present at a prominent DARPA symposium on Far Out Networking in Arlington, Va.
Looking forward, Krishnamachari said he feels excited about the potential of the next generation of wireless networks in cars, robots and beyond.
