I stumbled across this paper describing what could be an awesome app, but I'm not aware that anything has ever been made of it. Has anyone else heard of this or something similar? https://www.ideals.illinois.edu/bitstream/handle/2142/13916/paper.pdf
This paper is way better. They propose a system where people volunteer to transmit their OBDII fuel data and GPS data so that Fueoogle can log average fuel consumption for each segment of road. That data will then be used to calculate fuel efficient routes. Eco-route just guesses, but it doesn't know what streets have heavy traffic, a lot of stop signs, or crappy signal timing. Fueoogle would overcome these limitations. Over time, such an application could become very accurate.
I see - sounds like a great idea. The only issue I see is getting enough volunteers to have a good amount of data as most people are scared of being tracked even though it's anonymous.
Could be very easy for users of apps like Torque to volunteer to allow anonymized data to be uploaded to a Fueoogle server.
I was just playing with Torque and its map feature last night, was in awe once I noticed how you can map speed and other sensors right onto the map. I think Ford was working on a system like this tho.
Just before I get to work the most direct route is over a hill about 100 feet with a distance of 0.2 miles. I drive around it with a distance of 0.3 miles with no elevation change. None of the GPS auto software I have ever used takes me that route. But there is a lot of software that has good elevation information. Delorme is one that comes to mind. With the elevation data you would think that measurement data is unnecessary.
There's a lot more that affects mileage that elevation. Many cities jumped on the loop triggered signal fad and got rid of their synchronized lights so that now traffic lights are essentially random. With heavy traffic this creates the "red light to red light" syndrome on poorly planned streets. This kills mileage even more than a moderate amount of hill climbing. Collecting real-world data on a segment level would account for factors that are impossible to account for otherwise.
Thanks for that insight. Maybe that kind of information could to some extent be teased out of existing mobile phone tracking data and google traffic data. Road Traffic Monitoring and Provids Accurate cellular network Information How do Google traffic maps work? – The Blogs at HowStuffWorks
Plus, speed doesn't always tell you anything about fuel economy. That half-mile stretch of road with an average speed of 30 mph, is that a road with a 50 mph speed limit and a lot of red lights to accelerate from, or a road with a 30 mph speed limit? Fueoogle would operate much like Google's traffic data except it would be collecting fuel consumption data in addition to speed and location. When coupled with the vehicle make and model you can start to construct maps based on that data. For example, if a 2012 PiP burns 1 oz of fuel on a particular road segment that means it will do the same on similar segments.
