–
Tamir added that advanced driver-assistance systems (ADAS) reduce “the frequency of collisions but also their severity, which is particularly important to cyclists, who are one of the most vulnerable road users.”
We began our conversation at Ft. Meyer, the military base in Arlington, Va., where we were reviewing the feasibility of an automated shuttle service and discussing safety applications for ADAS in Defense Department vehicles housed on the base. I had some more questions for him about ADAS – a crucial precursor technology in the ongoing unveiling of automated vehicles (AVs).
Uri Tamir of Intel’s Mobileye
Kelley Coyner: While you’re working to solve that riddle of AVs being able to predict cyclist behavior, what are ways that AV tech can be used now to protect bicyclists and pedestrians?
Uri Tamir: There is no need to wait [for AVs to hit the road].
Mobileye’s systems provide sufficient time, either to the driver or to the vehicle’s actuation system, to take action and possibly avoid, or at least mitigate, the severity of a collision. Collision avoidance is a fairly immediate solution if you consider some other methods available to increasing bicycle safety (like bicycle helmets).
How do ADAS improve bike and pedestrian safety?
Each year, about 2 percent of motor-vehicle crash deaths are bicyclists. And that is after cities and the governments have spent millions of dollars on education and training programs to increase the use of safe-riding habits and helmet use.
ADAS does not replace those best driving/riding practices that governments are so desperate to promote – drivers should always be highly aware of vulnerable road users and cyclists should always ride with caution and make sure they wear helmets. But collision-avoidance systems that alert the driver of an imminent collision with bikes and pedestrians have the ability to make safer cities across the United States.
How many cars and trucks have you retrofitted with ADAS? Since the introduction of pedestrian and cyclist collision warnings, thousands of vehicles in the U.S. have been equipped with this life-saving feature as part of our ADAS aftermarket product. And thousands of new vehicles also have bicycle-detection technology powered by Mobileye. (Editor’s note: Take a look at all the other examples of what ADAS can accomplish besides spotting people walking and biking.)
Transit and sanitation fleets all over North America, including the New York Metropolitan Transportation Authority [MTA], which started equipping its buses earlier this year, work with us. And as part of the U.S. DOT Smart City Challenge, the winning city, Columbus, Ohio, will retrofit its entire fleet with Shield+, our ADAS solution for urban long vehicles.
What results have you seen so far?
Research recently released by the Transportation Research Board, the Washington State Transit Insurance Pool [WSTIP], and the University of Washington exploring the effectiveness of Shield+ on WSTIP’s assets was incredibly favorable about the positive impact that Shield+ can have on collisions and urban safety.
In that study, the buses logged more than 350,000 miles and close to 24,000 operating hours with Mobileye Shield. During that time, Shield+-equipped buses were involved in zero collisions with pedestrians or bicyclists. And they were not involved in any rear-end collisions.
Buses with Shield+ experienced 43 percent fewer pedestrian and cyclist collisions and blind-spot warnings. After WSTIP reviewed its insurance-claims history, it found that buses equipped with Shield+ showed a 58.5 percent potential reduction in vehicular and pedestrian claims due to collisions. That is a significant result, not only for safety but also for the [return-on-investment for] transit fleets.
Dish Network, the satellite television provider, analyzed more than 18 months and 12 million miles of Mobileye use, including collision rates and costs. It compared the data with its pre-Mobileye information and has seen a reduction of 85 percent in all types of collisions that Mobileye addresses and a 92 percent reduction in collision costs.
How much does an after-market retrofit cost?
Cost varies depending on the vehicle and the specific crash-avoidance system, but in this type of industry, the ROI is high. A transit authority could have retrofitted its entire fleet with collision avoidance for a tenth of the cost of a single jury verdict for a serious injury caused to a pedestrian by a right-turning bus.
Are you trying this on buses?
Buses are certainly a target for us, and there are unique challenges for the mass-transit industry when it comes to pedestrian and bicyclist safety.
The size and length of buses creates large blind spots on both sides of the vehicle and, with the bus’s A-pillar, this contributes to limited visibility for the bus operator. Greater turn radii pose a risk of a collision in almost every turn, particularly with a pedestrian or a cyclist hidden in one of the vehicle’s blind spots.
In addition, many transit buses drive on crowded city streets within close proximity to bicycle lanes.
Another challenge unique to the transit industry is approaching and leaving bus stops. When you think about it, transit buses are the only mode of transportation that requires a driver to move directly towards a group of standing pedestrians like they do when approaching a bus stop.
To solve these challenges, we created a multi-camera ADAS solution, up to four cameras on a regular 40-foot bus and six cameras for an articulated bus. The cameras communicate with each other and alert the driver both visually and through audible alerts. The alerts are directional, meaning, if the risk of a collision with a cyclist is detected from the right side of the bus, the bus operator will receive a visual and audible alert from the right-side display.
What is next in after-market applications? What is on your wish list?
One feature that I am particularly excited about is animal detection. Large animals crossing the road, such as deer or moose, pose significant collision risks in certain parts of the U.S., and there is a dire need for collision systems that can identify them.
