Convoys Against Congestion

Will Recker, G. Scott Samuelsen, and Michael McNally of UCI, stand with the electric and hybrid vehicles involved in ZEV-NET.

9.18.02 -- One key challenge of our time is to use technology to make life better - not worse. One place where technology could help has to do with using telecom and IT to facilitate communication among vehicles, leading to a drop in traffic congestion and improved driver safety. And Calit² is at the center of research and development on these issues.

California has more than 25 million motor vehicles - nearly two cars for every three people - enough to clog roads in most cities. And millions more cars are expected from immigration over the next decade. Given this situation, traffic is literally in a drag race with housing shortages and natural disasters to determine the winning obstacle to growth for the State's economy.

But the situation is even worse, according to Jim Wasserman of the Associated Press. In "Drivers push distractions from lap dogs to laptops," he writes, "In California, a state that has long defined American car culture, more commute-weary residents load up their interiors with VCR and DVD players, fax machines and dashboard video screens for satellite navigation systems. … Already, cell phones, in-car electronics and radio-CD systems represent the leading cause of inattention in crashes that killed 6,516 Californians and injured 413,913 last year."

These figures will likely climb with wider adoption of wireless messaging and broadband wireless, which can make television viewing a drive-time activity.

Isn't there a better way?

Calit² researchers are attacking the twin problems of traffic congestion and driver distraction in their two-campus, virtual-team approach. Prof. Will Recker, UCI Director of the Institute for Transportation Studies, and Prof. Mohan Trivedi, Director of the UCSD Computer Vision and Robotics Research Laboratory, are the Calit² layer leaders for Intelligent Transportation and Telematics. Together, they are leading development of the AUTONET living laboratory to test various concepts being studied Calit² researchers.

AUTONET, which builds upon significant support from Caltrans and other transportation agencies to develop and operate a real-world transportation laboratory in Orange County, refers to an autonomous, self-organizing communications and information system that encompasses roadways, stations, and vehicles, including the state of the driver of each car.

Complicating the traffic situation is a mandate by the State of California that by car model year 2003 - which begins now - at least 10% of new autos sold need to be electric, hybrid, or other low-emission vehicles. This challenge was turned into an opportunity by Recker and UCI colleagues, who created ZEV-NET (Zero Emission Vehicle Network), which uses a fleet of Toyota electric and hybrid vehicles to understand what role limited-range electric vehicles can play in California's commute mix.

In this context, one of the most interesting aspects of AUTONET goes beyond technology and traffic management to how technology interacts with social needs. AUTONET, with its pool of cars in ZEV-NET, is experimenting with a transportation modality that provides an interesting interface between our excessive reliance on personal cars and the public transportation infrastructure.

ZEV-NET focuses on enabling more people to commute via rail by providing cars at the rail station - so-called station cars - that multiple people can share to help "shave the peaks" off commute traffic caused by the one-person/one-car phenomena so common in California. Station cars would be made available at the train station for short-haul transportation in place of a car left overnight or a taxi.

Some commuters would love to take the train to get work done during their commute but are constrained from doing so because they need a car during the day. The quid pro quo of station cars is that if you took the train and left your car at home, a station car could be made available to you, via dynamic scheduling available on your wireless PDA wherever you were, as long as another station car user was able to leave the car in your vicinity. The plan is for the vehicles themselves to communicate in a peer-to-peer fashion for management, control, and other purposes, such as keeping the reservation system informed.

The "net" in ZEV-NET comes from applying Calit² expertise in communications and IT to create what some might call the World-Wide Wagon Train, since it involves establishing a flow of conversation - or its car-based equivalent - among vehicles. The implications of this work for drivers could be as significant as the World-Wide Web has proven for surfers.

"What if, as you drove, you could get messages from cars that were coming from where you were headed?" says Recker who just turned 60 but has the enthusiasm of a teenager ready for his first car. "You could learn of changing traffic conditions. Your next-generation navigation system could interpret these and advise alternate routes. In case of an accident, cars passing by the scene could pick up the signal and broadcast it on to the emergency responders, while cars following behind could get the message to slow down or get off at earlier exits, keeping the road more clear for the emergency vehicles."

Visitors to the control center at UCI's Institute of Transportation Studies see computer graphics showing moving 3-D blocks of color where cars are, as well as real-time video angled to look for traffic jams. Currently the data is generated by passive sensors in the pavement and the state of control lights at freeway entrances. With AUTONET, this control room will have vastly more - and more finely grained - data with all the information relayed by each "wagon" in the collective "train."

Recker is a good man to lead the wagon train. He has teamed with Calit² researchers Rajesh Gupta and Yehuda Bock to apply core Calit² technologies, including wireless communications, sensornets, geolocation, and, in the near future, custom-designed chips to improve the quality and quantity of car-to-car communication.

Gupta, who recently moved from UCI to UCSD, is working with Compaq iPAQ PDAs that send and receive navigation and related information while hooked into the transportation information infrastructure. Gupta studies mobile computing systems and their implementations as integrated Systems-On-a-Chip (SOC). SOCs hold the promise of increased functionality at drastically reduced system costs that will make it far more likely for auto companies, which install anywhere from 5 to 80 chips in a car, to include communications, networking, and entertainment as options.

Bock, a research geodesist based at the Cecil H. and Ida M. Green Institute of Geophysics and Planetary Physics (IGPP) at UCSD and Director of the Scripps Orbit and Permanent Array Center (SOPAC), is bringing the geolocation abilities of GPS (Global Positioning System) satellites to work for AUTONET. Modifying the methods he developed for monitoring seismic motion of the solid Earth, Bock believes that, through more sophisticated GPS algorithms, vehicles can have their positions identified not just within 100 yards, as with today's navigation systems, but closer to 10 meters. With further refinement, GPS would be in a position to pinpoint more exactly what lane a car is in, which could then be used by guidance systems to advise the driver when to change lanes, in what direction, and so forth based on current traffic conditions.

All this work is complemented by projects led by professors Mike McNally and Scott Samuelsen of UCI, and Mohan Trivedi. McNally and his team of graduate students at UCI have designed and installed small communication devices that relay real-time status information about the location and usage of vehicles in the program, as well as "intelligent" online surveys that shed light on changes in driver behavior accompanying these new capabilities.

Samuelsen and his colleagues at the National Fuel Cell Research Center at UCI are pioneering the concept of Power Parks, a distributed power analog to the distributed information concept of AUTONET, as a clean, fuel-cell alternative for generating the power for clean running ZEV-NET electric vehicles, without simply transferring the pollution production to distant electricity generating plants.

Prof. Trivedi at UCSD studies driver distraction and using robots to monitor traffic from fixed locations, a set of technologies so futuristic that I will return to it as the subject of a future article, as I will for the research of professors Gupta and Bock.
Because consumer demand - and willingness to pay - for any of the technologies being investigated here will not be clear for a while, the technologies may appear as after-market, add-on gizmos rather than built in to new cars. This, in turn, could point to a new sort of consumerist "living lab" as drivers sort out the technologies of choice and begin to determine what we will see at Pep Boys in the next few years.

Might not it be interesting if, say, consumer electronics, post-purchase, could themselves be used to design the next generation of the same products? Or to guide the information content associated with those products? AUTONET may well provide the foundation to begin making that possible.

----------------------------------------------------------------------

1. http://www.auto.com, August 23, 2002.
2. According to the Wireless World Forum, by 2007, 20% and 12% of new vehicles will be Bluetooth- and 802.11-enabled, respectively.
  3. ZEV-NET, at the State level, refers to Zero Emission Vehicle Network Enables Transport; see http://www.zevnet.org.
4. Historian Daniel Boorstin uses the wagon train as Exhibit A in The Americans: The National Experience to prosecute his case that community, not individualism, was and is at the heart of America's success. He points out that wagon trains were able to cross some of the hottest, coldest, steepest, muddiest terrain humans ever faced by combining the insights, information, and ingenuity of fellow travelers.
5. The traffic signals receive data from passive sensors, but they do not generate data per se.
6. Keep in mind, though, that this data is being used for research purposes rather than to control the traffic per se.