In the past, I was into hypermiling. I have a 2002 Volkswagen Jetta TDI that I purchased back in 2012. The previous owner treated it very roughly and it needed some good maintenance to get it back to proper performance. I replaced the tires with a Goodyear low rolling resistence model, the windshield wipers with a lower drag model, and got the turbo and vacuum system in order. Once those things were taken care of, I started working on my driving technique. I had a 60 miles daily roundtrip for work and learned my route very well, including inclines, declines, and the best places to accelerate and decelerate. I was even once able to get a 60 MPG average on a tank of fuel. I became very tuned to how different choices when driving impacted my fuel economy.
The modern smartphone is a computational powerhouse. We have GPS, accelerometers, and barometers. Here is my idea, free to anyone that implements it (if you do, please just let me know and give me a crack at testing it): sense the various factors of a commute and crunch the sampled data over time to make recommendations to drivers as when to accelerate, coast, decelerate, what speed to maintain, etc, all in an effort to improve energy efficiency of driving. This could take into account different drivetrains and maybe even mature to a point where there are recommendations for specific cars and roads that are effectively crowdsourced. For instance, a gasoline car will have a different acceleration profile compared to a diesel, and any vehicle with regenerative braking will have different characteristics to improve fuel efficiency when it comes to decelerating. In addition, the app could connect to a bluetooth ODBII interface attached to the car to utilize the reporting mechanism from the ECU of the vehicle and its various sensors (incoming air temp, engine temp, engine load, etc).
ON ANOTHER NOTE: Related to electric cars – driving an electric car is a very different experience. You push the pedal both to accelerate and maintain speed, similarly to a traditional vehicle, but the braking component of coasting is more marked than a gasoline/diesel car. Effectively, you [regeneratively] brake by letting off of the accelerator pedal; you can basically drive by using only one pedal (pressing the brake pedal engages the mechanical brakes and is still required for situations where you require more braking force). A logic mechanism should exist that allows an electric car to coast without regenerating resistence. While likely to be unused by most drivers, this would allow the vehicle to be the most efficient as possible when coasting if you aren’t intending to slow down [drastically]. Yes, regenerative braking will recapture some energy, but it is more efficient to just coast without resistence in some cases. The process of regenerative braking is a great boost, but you do not recapture all of the energy… it is just impossible given the reality of physics.
I am not an app developer and I haven’t yet reached a tipping point of motivation to get there. Perhaps one day I will and I will take up the mantle of making such an app. However, feel free to take this idea and run with it if you have the skills and desire.