Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Smartcar - automated climate control and charge scheduling for Model S

This site may earn commission on affiliate links.
Maybe if the software could figure out what time my alarm went off, and how many times I hit snooze. Then it might have a shot of getting my morning strategy correct.

Hey! I'm the guy behind Smartcar, I was curious what you thought of this: We actually felt it might be even nicer to let you link your Google Calendar so that we can determine the times of upcoming events and their respective locations.

P.S. An alarm clock would work too as we enter the age of everything being connected, but the number of folks with an alarm which packs an API might be a bit limited today. :)
 
Yea this guy is charging for something that is brain dead easy to program using the same rest APIs. Honestly this is mostly hype there's not really any value added by what he's offering.

Thanks for your feedback. We surely are doing a bit more than a simple REST API call. For starters, we've have a Data Scientist who's working on some incredible machine learning to help learn your driving patterns.

What this means:
1) It learns what temperature you like at what time, even if your schedule varies heavily.
2) It learns how long it takes to bring your car to the right temperature.
3) It learns how much it costs and how much power it takes to do that.

We of course have taken measures to ensure the product 1) still lets you have control 2) is as secure as possible. In my opinion, technologies like this are why people love products like the Nest Thermostat or Google Now so much. It's very tough to build and takes iterations to get it right.

We also want to make things easier for those who don't want to use learning features:
1) You can easily set up an advance schedule to heat or cool your car at a desired time each day. You won't have to write any code, maintain it, secure it, or leave a computer/server running. And you can do it from a friendly interface available on smartphones, tablets, and desktop.
2) When it comes to charging, we found that electric rates vary between weekdays and weekends. These hours also change between the summer, winter, and even on holidays. We continuously fetch and update this data for all the major electric providers to ensure your car is being charged at the best time to minimize your bill.

I'd be more than happy to answer any more questions you may have. I'm open to criticism as it's very necessary to get the product crafted properly.
 
In WA the rates are fixed so rate optimization is of little value. Also given that a high SOC is undesirable you would seem to have competing goals of price rate optimization vs. as-late-as-possible charge completion to preserve long-term battery health. Also would be curious to know how you plan to manage the target SOC. If you employ an as-little-as-needed approach you run the risk of having to pay peak electrical rates to pick up a boost charge during the day.
 
one thing I need in short term is ability to set number of miles I need for weekend and weekdays. This will avoid me taking guess (how long to charge, how much charge level to set) and remembering to charge on different times (I use PG&E and E6 plan). I see that you support this already. If this can also take care of charging during non-peak hours, that will be very helpful.

It is also important to have controls to choose what is important for individual owners (as it is not possible to take one-size-fits-all approach as @ACDriveMotor pointed). For me, 1. number of miles at a given time 2. abstract PG&E complexities (as recommended hours change on so many factors). What this means is, I am ready to pay few additional cents for comfort. But I am pretty sure people have different priorities (just like for ACDriveMotor it seems like Battery health is more important).
 
@sahaskatta I love the idea. Having nest thermostats, and making heavy use of IFTTT for lots of "zero clicks" interactions I want to say again this is almost "to be expected" from Tesla & ecosystem. Good execution around how the system behaves and interacts and learns from type I and type II error situations is key, but you won't get there if you don't get started. Letting people express conflicting priorities implicitly (by choosing when relevant) or explicitly (preferences/config) may help (eg save battery cycles over comfort, expressed as a slider bar). Nest's model of adjusting an explicit schedule with implicit behaviors is great. And integrating with ifft would allow you to incorporate more events into the model- like a belkin motion sensor in the bathroom triggering the heating to happen for 45 m from now.
 
Thanks for your feedback. We surely are doing a bit more than a simple REST API call. For starters, we've have a Data Scientist who's working on some incredible machine learning to help learn your driving patterns.

What this means:
1) It learns what temperature you like at what time, even if your schedule varies heavily.
2) It learns how long it takes to bring your car to the right temperature.
3) It learns how much it costs and how much power it takes to do that.

We of course have taken measures to ensure the product 1) still lets you have control 2) is as secure as possible. In my opinion, technologies like this are why people love products like the Nest Thermostat or Google Now so much. It's very tough to build and takes iterations to get it right.

We also want to make things easier for those who don't want to use learning features:
1) You can easily set up an advance schedule to heat or cool your car at a desired time each day. You won't have to write any code, maintain it, secure it, or leave a computer/server running. And you can do it from a friendly interface available on smartphones, tablets, and desktop.
2) When it comes to charging, we found that electric rates vary between weekdays and weekends. These hours also change between the summer, winter, and even on holidays. We continuously fetch and update this data for all the major electric providers to ensure your car is being charged at the best time to minimize your bill.

I'd be more than happy to answer any more questions you may have. I'm open to criticism as it's very necessary to get the product crafted properly.

I think this is pretty cool, although I'm not sold yet :). Would be great to be able to sign up for updates at your website. For me given that I do not have a regular schedule, curious to see how the learning part works. Also would be great to have this plug into my home automation setup.
 
most of this is already available via open source VisibleTesla and it probably wouldn't take but a few hours of programming to extend it with some simple 'machine learning' algorithms to accomplish all of the above. Plus that's more secure as putting this up on the cloud potentially leave's everyone's Tesla credentials vulnerable if hacked.