show details 8:59 AM (2 hours ago) |
Ford adds fuel-efficient routing to Sync
My Ford Touch screen displays navigation data and offers choice of routes that, in addition to ones for shortest distance and fastest route, are most fuel-efficient.While current onboard navigation systems can provide a driver with the shortest and fastest routes to a destination, a new one can tell the way to get there while using the least amount of fuel. TeleNav Inc., the route-mapping provider for the new system, available through theFord Sync system with the My Ford Touch screen, has developed a sophisticated algorithm to produce that level of route guidance.
The algorithm takes a lot of computing power, but because the navigation system resides off board in cloud servers, it can continuously be updated with road changes and conditions. Further, it can accumulate other data, including some unavailable to the onboard navigation system, and do more with it all.
The company worked with Ford-supplied fuel consumption data, not only numbers for all steady speeds but also curves plotted for all other operating conditions monitored by the algorithm. The fuel-economy algorithm considers the following, explained TeleNav automotive business development manager James Grace:
• Speed limits for the entire route. The fuel economy and all three (shortest, fastest, most-efficient) time-to-destination calculations are based on the driver following the speed limits.
• Number of stops required by stop signs and traffic lights. With the latter, a probability factor is used to estimate number of times the car will "make" the light vs. having to stop for it.
• Amount of vehicle-stopped idling time. In this case there is a probability-based estimate of the time stopped at each traffic light. The algorithm uses a varying time frame for traffic light stops, from 20 s to about 2 min, Grace said, to better simulate real-world events. If a car were a hybrid equipped with engine stop-start, the algorithm would be adjusted to correct for this, Grace added.
• Traffic on the routes, both historical data and current, real-time traffic conditions. If necessary, the algorithm inserts a correction factor into the calculations for average speed that would likely be achieved.
The speeds are tied to the distances each speed limit is posted. The algorithm assumes moderate braking to each stop and a moderate acceleration from zero to the posted speed. The algorithm calculates the extra fuel consumed for both the brake-to-stop and acceleration-from-stop as well as fuel consumption during the idle stop itself.
Although the algorithm assumes all the roads are flat, Grace said, it does consider road curvature, and if the curvature is 30% or greater from straight ahead, it assumes the driver will apply the brakes to slow down and then accelerate to the posted speed. It adjusts fuel consumption computations for each road curve-induced event the vehicle would encounter.
The Ford data on fuel consumption is vehicle-specific, and if there is a significant difference from one calibration to another, TeleNav uses both, picking the specific one, such as turbo vs. nonturbo.
Although shorter routes—perhaps 11-12 mi (18-19 km)—show differences, they may not be dramatic. However, TeleNav produced an example for AEI that did.
With a start point in Nashville, TN, on a weekday afternoon, it calculated routes to a landmark barbecue restaurant in Memphis, a distance of over 200 mi (320 km). The shortest (and in this case, also the fastest) route was 204 mi (326 km), with a travel time of 3 h, 30 min. The most economical route was actually much longer at 228 mi (365 km), with a travel time of 5 h, 11 min.
Is it really possible that the much longer route actually will result in less fuel used, not just higher mileage? Why wouldn't the faster route, which is primarily highway, result in less fuel used?
The answer: although the faster route had an average speed of just over 58 mph (93 km/h), it was primarily at highway speeds of 75 mph (120 km/h). This compared with an average of 44 mph (70 km/h) for the economy route, with most mileage traveled at 50-55 mph (80-88 km/h), Grace said. The Ford fuel economy data for that vehicle calibration included 23 mpg for 75 mph and 35 mpg for 50 mph (over 50% higher), so lower speeds more than compensated for 12% greater distance.
TeleNav recognizes that people do not travel at exactly the speed limits, even if traffic is not a factor. They may go slower on high-speed highways, for example, and get better mileage than with the predicted eco-route, Grace conceded.
However, TeleNav ran a series of validation routes, using a closer-to-normal driving technique, in the San Francisco, CA, area. The testing, Grace said, confirmed that the eco-route was consistently better than the other routes. The tests were run on a fill-up to fill-up basis at the same gas stations, he said. Additional tests will be run to provide greater accuracy with the comparative fill-ups and provide a statistically valid number for the average improvement in fuel economy.
There is obviously room for improvement in the algorithm, Grace said, including use of available topographical data, to compensate for driving through hilly areas. But TeleNav and Ford are satisfied with the initial indications, he added.
The algorithm takes a lot of computing power, but because the navigation system resides off board in cloud servers, it can continuously be updated with road changes and conditions. Further, it can accumulate other data, including some unavailable to the onboard navigation system, and do more with it all.
The company worked with Ford-supplied fuel consumption data, not only numbers for all steady speeds but also curves plotted for all other operating conditions monitored by the algorithm. The fuel-economy algorithm considers the following, explained TeleNav automotive business development manager James Grace:
• Speed limits for the entire route. The fuel economy and all three (shortest, fastest, most-efficient) time-to-destination calculations are based on the driver following the speed limits.
• Number of stops required by stop signs and traffic lights. With the latter, a probability factor is used to estimate number of times the car will "make" the light vs. having to stop for it.
• Amount of vehicle-stopped idling time. In this case there is a probability-based estimate of the time stopped at each traffic light. The algorithm uses a varying time frame for traffic light stops, from 20 s to about 2 min, Grace said, to better simulate real-world events. If a car were a hybrid equipped with engine stop-start, the algorithm would be adjusted to correct for this, Grace added.
• Traffic on the routes, both historical data and current, real-time traffic conditions. If necessary, the algorithm inserts a correction factor into the calculations for average speed that would likely be achieved.
The speeds are tied to the distances each speed limit is posted. The algorithm assumes moderate braking to each stop and a moderate acceleration from zero to the posted speed. The algorithm calculates the extra fuel consumed for both the brake-to-stop and acceleration-from-stop as well as fuel consumption during the idle stop itself.
Although the algorithm assumes all the roads are flat, Grace said, it does consider road curvature, and if the curvature is 30% or greater from straight ahead, it assumes the driver will apply the brakes to slow down and then accelerate to the posted speed. It adjusts fuel consumption computations for each road curve-induced event the vehicle would encounter.
The Ford data on fuel consumption is vehicle-specific, and if there is a significant difference from one calibration to another, TeleNav uses both, picking the specific one, such as turbo vs. nonturbo.
Although shorter routes—perhaps 11-12 mi (18-19 km)—show differences, they may not be dramatic. However, TeleNav produced an example for AEI that did.
With a start point in Nashville, TN, on a weekday afternoon, it calculated routes to a landmark barbecue restaurant in Memphis, a distance of over 200 mi (320 km). The shortest (and in this case, also the fastest) route was 204 mi (326 km), with a travel time of 3 h, 30 min. The most economical route was actually much longer at 228 mi (365 km), with a travel time of 5 h, 11 min.
Is it really possible that the much longer route actually will result in less fuel used, not just higher mileage? Why wouldn't the faster route, which is primarily highway, result in less fuel used?
The answer: although the faster route had an average speed of just over 58 mph (93 km/h), it was primarily at highway speeds of 75 mph (120 km/h). This compared with an average of 44 mph (70 km/h) for the economy route, with most mileage traveled at 50-55 mph (80-88 km/h), Grace said. The Ford fuel economy data for that vehicle calibration included 23 mpg for 75 mph and 35 mpg for 50 mph (over 50% higher), so lower speeds more than compensated for 12% greater distance.
TeleNav recognizes that people do not travel at exactly the speed limits, even if traffic is not a factor. They may go slower on high-speed highways, for example, and get better mileage than with the predicted eco-route, Grace conceded.
However, TeleNav ran a series of validation routes, using a closer-to-normal driving technique, in the San Francisco, CA, area. The testing, Grace said, confirmed that the eco-route was consistently better than the other routes. The tests were run on a fill-up to fill-up basis at the same gas stations, he said. Additional tests will be run to provide greater accuracy with the comparative fill-ups and provide a statistically valid number for the average improvement in fuel economy.
There is obviously room for improvement in the algorithm, Grace said, including use of available topographical data, to compensate for driving through hilly areas. But TeleNav and Ford are satisfied with the initial indications, he added.
No comments:
Post a Comment