Thank you so much for your response. These are great questions. Hope my answers below provide some insight.
1) 200,000 mobile weather stations--how do these function?
We have mobile weather stations connected to vehicles that are driving lots of routes across the U.S. There are not 200,000 weather stations, we have collected data from our weather stations in over 200,000 locations so far. This data has been collected over the last 8 years and we have been studying the effects of topology and geographic features on weather conditions at the street level. It has been my life’s work in professional cycling, and was my first job in the Tour de France: driving the course ahead of the race (going ‘avant course’ as they say in French) and reporting back the conditions to the team. This information was used to adjust the team's pacing strategy and equipment selection on the day.
We have applied a similar methodology to vehicles and now have a team devoted to this project. We study the effects that various structures have on wind, temperature, etc. in different regions, and suggest route selection and behavioral changes while traversing a given route. We continue to improve on our measurement device that we call, the Scout, and have developed a set of sensors that is capable of acquiring precise conditions of the environment at the road surface. I have put together a PDF that gives more detail with pictures of the Scout, and will post it to this thread at the end of the weekend. The goal is to provide our data in a mobile app to eliminate the need for a hardware device mounted on vehicles.
What's the coverage like in any given area? How up-to-date is the app; if I'm driving in the middle of nowhere in southern Illinois, is this going to be accurate?
Because our technology relies on knowledge of the data along your specific route, roads that we have not traveled will include data from the National Weather Service. So the information we provide is at least as accurate as the weather station data that you would look up online. This data is then put through our algorithms from what we have learned from the topology in similar regions and adapted to your route. So the NWS data is adjusted to the geographic features. We continue to collect more data and provide the level of accuracy in different regions. As this information becomes available, we will release it on our website.
Does the app update conditions in realtime?
The app updates by making an API call to our server in realtime. We reduce the data consumption on your phone by optimizing the API calls based on the data collected in that area.
2) It seems to me that the two weather parameters that will affect fuel economy the most are wind and temperature. These don't vary much over even a large area of the country on any given day. What benefit will I derive from the app that I couldn't get from checking weather conditions online before I leave on a trip and adjusting my route accordingly myself?
You may have pulled this together from my answers above, but our data is different than that from a weather station due to being measured at the street level. Topology and geographic features impact the wind and temperature (and other metrics), and our dataset reflects that. Weather stations can be located many miles away from your vehicle and far from the surface of the earth. And because of that, the data is quite different than what we provide.
Separately, if we have not collected data in your area and the app is providing data from NWS, then the app at least provides the convenience of looking up the weather conditions automatically and planning an optimal route and strategy for you.
3) On a similar note, if I'm traveling between two fixed points, how much are the details of my route going to matter when I'll have to travel roughly the same distance against a headwind, or with a crosswind, or a tailwind--i.e. in the same general conditions? Do you have any metrics yet on how much fuel this could save?
If you are traveling between two fixed points, and for simplicity, let’s say that you are going in a straight line with a constant headwind and no interference from other objects (i.e. traffic, buildings, trees, etc.). The first potential benefit is to tell you what time of day (or date if you have that option) is best to travel this route. If your time is not flexible, then we can also tell you how much money can be saved by sacrificing some time from adjusting your speed (in a tailwind route, this would not be the case).
In a more complicated example where you adjust your heading along your route, aerodynamics is going to change due to the effective headwind. And pacing may change given this scenario. Fuel economy can be optimized. You will notice that we have provided three different categories of vehicles for you to choose from. As we research more in wind tunnels, using CFD, and more field experiments, we will provide more vehicle options. Also, the geographic features along a route can change the direction of the wind. So you can benefit from this knowledge.
Now, let’s complicate it some more by including geographic features such as buildings, trees, hills, etc. In these areas, we can tell you where aerodynamics has the largest affect and adjust your speed accordingly to optimize your fuel efficiency. Our calculations include traffic, so in some circumstances where traffic is congested in isolated sections of your route, the time penalty of making these speed adaptations is minimal.
The Scout collects the following data and we will be integrating all of it into the app as it becomes available:
-Wind speed
-Wind direction
-Wind temperature
-Temperature
-Humidity
-Pressure
-Precipitation
-Road gradient
-Surface roughness
-Surface condition (wet, dry, ice)
-Corner radius
We are still developing more features to include in this app and would appreciate any interested users to provide their feedback. It is free, so please send us an email at robby<at>winningAlgorithms.com if you want to give it a go.
Thanks again for the questions. I am working on a more detailed doc and we are updating our website in the next couple of weeks. Please check it out then for more details.
Best,
Robby