Chevy Bolt EV Forum banner

1 - 20 of 67 Posts

·
Registered
Joined
·
33 Posts
Discussion Starter #1
A few months ago, when I got my Bolt, I was frustrated by the lack of good planning tools for the Bolt. The OnStar app has a route planner, but there isn't a real good way to use that in the car, and it is woefully inaccurate around me, directing me to charge at Level 2 stations for 9 hours instead of picking up fast chargers just slightly off my route.

Long story short, the creator of A Better Route Planner (Bo) was gracious enough to allow me to contribute code to add Bolt EV Route Planning to his tool, and now it's live!

Try it out and see if you find it useful!

If you have any issues, feel free to contact me directly or post in this thread.

Links:
A Better Route Planner
Release Notes
 

·
Registered
Joined
·
63 Posts
Seems really good.

I have a few observations where there might be room for improvement when planning a trip in a Bolt EV:

1) It seems to like to plan a lot of short stops when possible to keep charging rates high and charging times low. I'm not sure if that's a good thing or not.

2) In the US there's still a lot of places where you'll need to stop at a level 2 to get where you're going. While it will happily plan a coast to coast trip in a Tesla, if you try it in a Bolt, even if you manually tell it to stop at some L2 chargers to fill in the gaps, it either tells you there's no route, or gives you a routing error.

3) After a route is done, clicking on the link to open the route in Google Maps doesn't always result in the same route. The waypoints (chargers) are the same, but the route taken isn't. Example:

Since the forum wont let me post links, just plan a route from the charger in Elmira, New York to any charger just over the border in Canada at Niagara Falls. ABRP has the route leaving Elmira on state road 96, while if you click on the link to OPEN the route in Google Maps, the route shown is on state road 89. 89 is the road the runs along the west side of one of the Finger Lakes, Lake Cayuga, while 96 is a parallel road a few miles to the west. This is a 0-stop route.

If you switch the vehicle to a Tesla, the same thing happens. ABRP starts you on 96 while opening the route on Google Maps starts you on 89.
 

·
Registered
Joined
·
215 Posts
Thanks for the update. Looks really interesting.
Does this exist as an app for iPhone or Android?
It didn't work in Explorer, but does in Chrome.
Is there a way for users to input charging stations, or better yet for the network to mine PlugShare and other sites?
 

·
Registered
Joined
·
33 Posts
Discussion Starter #4
Seems really good.

I have a few observations where there might be room for improvement when planning a trip in a Bolt EV:

1) It seems to like to plan a lot of short stops when possible to keep charging rates high and charging times low. I'm not sure if that's a good thing or not.

2) In the US there's still a lot of places where you'll need to stop at a level 2 to get where you're going. While it will happily plan a coast to coast trip in a Tesla, if you try it in a Bolt, even if you manually tell it to stop at some L2 chargers to fill in the gaps, it either tells you there's no route, or gives you a routing error.

3) After a route is done, clicking on the link to open the route in Google Maps doesn't always result in the same route. The waypoints (chargers) are the same, but the route taken isn't. Example:

Since the forum wont let me post links, just plan a route from the charger in Elmira, New York to any charger just over the border in Canada at Niagara Falls. ABRP has the route leaving Elmira on state road 96, while if you click on the link to OPEN the route in Google Maps, the route shown is on state road 89. 89 is the road the runs along the west side of one of the Finger Lakes, Lake Cayuga, while 96 is a parallel road a few miles to the west. This is a 0-stop route.

If you switch the vehicle to a Tesla, the same thing happens. ABRP starts you on 96 while opening the route on Google Maps starts you on 89.
Thanks for the feedback!
On 1, yeah, we optimize to minimize total trip time. You can adjust this by increasing your charger overhead time ("Time to open charge port" in the settings). The longer overhead at each charger, it'll try to minimize your time at each charger. Another good way is to pick a charger as a "long break" (click on the charger, select "Take a long break here") and it'll assume you're going to charge to full there, and will give you fewer stops after that.

On 2, agreed. That's on my to-do list, I don't really want to have it route you through Level 2 chargers as that's not realistic for a trip. But I do want it to provide you a better idea of where it failed in route planning, and let you fill in the gaps if you know of places to charge overnight.

On 3, That's an unfortunate side effect of the way the routing is calculated. We use OSRM (Open Source Routing Machine) to generate our routes. This is of course a different routing engine from Google's, and will at times pick different routes to the same goal. All we do is send Google the list of waypoints, and let it determine its own route, and the API to do so doesn't allow us to specify a route. In theory, if money were no object, we could use Google's directions engine on both ends, then we'd have no differences, however those services of Google are pretty expensive. I'll see if there's anything we can do in our inputs to OSRM, but I'm not sure how much control we have on that.

The danger is, of course, that if you don't realize this has happened, Google takes you on a longer route than was originally calculated, and uses up more energy than we planned for. I'll have to give this one some thought.

Again, excellent feedback!

Thanks for the update. Looks really interesting.
Does this exist as an app for iPhone or Android?
It didn't work in Explorer, but does in Chrome.
Is there a way for users to input charging stations, or better yet for the network to mine PlugShare and other sites?
We don't have an app yet, but it is on our radar. It would be a lot of work, but valuable. Before I joined the team, it was just Bo, and now there's still just the two of us, we'll see how much bandwidth that adds to the development team.

I have a known issue in Explorer I need to fix tonight after work, should be resolved soon (doesn't support one of the built-in javascript functions I was using, will have to do a small rewrite.)

Your last question is one of the best ways you can help us out! I wish we could use PlugShare, but they don't allow access by non-commercial / non-profit projects, even if we were willing to pay. For the moment, our biggest US database is OpenChargeMap, which has an admittedly difficult-to-use interface, but their API is actually accessible. For now the best way to improve our database is to submit edits to OCM.

The other way we could use help is in collecting good driving data to improve the models (driving and charging). If you've got an OBD reader, android phone, and are willing to send me your power consumption data, shoot me an email and I'll help you get set up.
 

·
Registered
Joined
·
63 Posts
Your last question is one of the best ways you can help us out! I wish we could use PlugShare, but they don't allow access by non-commercial / non-profit projects, even if we were willing to pay. For the moment, our biggest US database is OpenChargeMap, which has an admittedly difficult-to-use interface, but their API is actually accessible. For now the best way to improve our database is to submit edits to OCM.

The other way we could use help is in collecting good driving data to improve the models (driving and charging). If you've got an OBD reader, android phone, and are willing to send me your power consumption data, shoot me an email and I'll help you get set up.
It might not be a bad idea to include both of those suggestions, with appropriate background information, somewhere within ABRP itself. For example, a link that says "Add a missing charger" could open a help page that directs users to OpenChargeMap.
 

·
Registered
Joined
·
33 Posts
Discussion Starter #7
It might not be a bad idea to include both of those suggestions, with appropriate background information, somewhere within ABRP itself. For example, a link that says "Add a missing charger" could open a help page that directs users to OpenChargeMap.
That's a great idea, I'll talk to Bo about how we'd want to implement that. One thing we want to avoid is becoming a charger database ourselves, there's enough of those out there already.
 

·
Registered
Joined
·
63 Posts
Speaking of L2 chargers, when I did add them as waypoints, the waypoint display had four parameter settings for each of the L2 waypoints. I couldn't figure out exactly what each one was used for. How are these used?

Another thought -- would it be possible to have the "more settings" section (temperature, wind, road condition, and extra weight) expose how much they affect vehicle range? If ABRP has decent models for those, it would be nice to know how much this affects efficiency. It could make everyone a better driver.

Some more ideas: Settings for having AC, heat, or seat-warmer plus steering wheel heater turned on.

Finally: Showing the vehicle range at various speeds (with all the various settings) would be informative. Certainly, this must be calculated before doing any of the routing.

Hmmmm. One more. "Reference Speed" is relative to the speed limit, right? If it's set to 110%, and the speed limit is 60, the planner assumes you'll be doing 66? If so, I'd suggest doing it differently. I think it would be more intuitive to have "Speed Limit Offset". 0 would meen doing the speed limit. +10 would be 10 miles (or km) over the speed limit. -5 would be 5 miles below the speed limit. You might want two settings; one for highways and one for local roads. It's just a thought.

Of course, feel free to ignore all of this.
 

·
Registered
Joined
·
1,105 Posts
The other way we could use help is in collecting good driving data to improve the models (driving and charging). If you've got an OBD reader, android phone, and are willing to send me your power consumption data, shoot me an email and I'll help you get set up.
I just sent you a PM about this.

Questions:
1. Does it take into account elevation changes? I used it to plan a route almost exactly 200 miles away, there are two viable routes to get there with very similar distances (within a couple miles) but ABRP chose the route with significantly more elevation changes. The route it took IS the more popular one with more time on the Interstate and higher average speed limits.
2. Does it build any buffer time around charging stops, or does a 1 minute charging stop (yes it suggested one for this trip) add exactly 1 minute to the trip?
3. Related to above would be nice if you could build in a setting where it advised you if you could save time overall by avoiding a charging stop by driving a little slower. I know you can check the "slower if necessary" option but that involves re-running the trip.

I do think options for AC/Heater usage are pretty important to add, as those could add significantly to consumption.
 

·
Registered
Joined
·
63 Posts
2. Does it build any buffer time around charging stops, or does a 1 minute charging stop (yes it suggested one for this trip) add exactly 1 minute to the trip?

In the waypoints window, click on "Show Settings", and then click on "Show More Settings". "Time to Open Charge Port" is the setting you're looking for. That's the extra time it adds to each charging stop.
 

·
Registered
Joined
·
33 Posts
Discussion Starter #11 (Edited)
Speaking of L2 chargers, when I did add them as waypoints, the waypoint display had four parameter settings for each of the L2 waypoints. I couldn't figure out exactly what each one was used for. How are these used?
This exact issue is on my to-do list, it's not very clear what those fields mean. For now, though, here's what they do (from the left):
Arrival % = minimum arrival percentage to the waypoint
Power (kW) = power provided by the location
Charge time = Time you plan to charge at the charger
Charge % = how full you want to be when you leave the charger.

Charge time and Charge % are nearly interchangeable (though if you pick charge %, it will try to minimize your time at that location). You can leave all of the fields blank, and it does different things depending on what's input.

Another thought -- would it be possible to have the "more settings" section (temperature, wind, road condition, and extra weight) expose how much they affect vehicle range? If ABRP has decent models for those, it would be nice to know how much this affects efficiency. It could make everyone a better driver.

Some more ideas: Settings for having AC, heat, or seat-warmer plus steering wheel heater turned on.

Finally: Showing the vehicle range at various speeds (with all the various settings) would be informative. Certainly, this must be calculated before doing any of the routing.

Hmmmm. One more. "Reference Speed" is relative to the speed limit, right? If it's set to 110%, and the speed limit is 60, the planner assumes you'll be doing 66? If so, I'd suggest doing it differently. I think it would be more intuitive to have "Speed Limit Offset". 0 would meen doing the speed limit. +10 would be 10 miles (or km) over the speed limit. -5 would be 5 miles below the speed limit. You might want two settings; one for highways and one for local roads. It's just a thought.

Of course, feel free to ignore all of this.
I really like the "show estimated range" idea, I'll look at implementing it at some point, but it'll probably be a little lower on my priority list. Thing about reference speed is, you're not going to drive 10mph above a 30mph speed limit (I hope) that's a lot more of a difference than 10mph above a 60mph speed limit. Could implement a calculator to show you what your reference speed means at 65mph though, to give you a gauge.

I just sent you a PM about this.

Questions:
1. Does it take into account elevation changes? I used it to plan a route almost exactly 200 miles away, there are two viable routes to get there with very similar distances (within a couple miles) but ABRP chose the route with significantly more elevation changes. The route it took IS the more popular one with more time on the Interstate and higher average speed limits.
2. Does it build any buffer time around charging stops, or does a 1 minute charging stop (yes it suggested one for this trip) add exactly 1 minute to the trip?
3. Related to above would be nice if you could build in a setting where it advised you if you could save time overall by avoiding a charging stop by driving a little slower. I know you can check the "slower if necessary" option but that involves re-running the trip.

I do think options for AC/Heater usage are pretty important to add, as those could add significantly to consumption.
1 - Yes, fairly conservatively, actually, we account for power in going uphill, and only assume you get relatively little back going downhill (~30% if memory serves).
2 - Answered below, you can set that yourself to whatever you end up using.
3 - True, we already provide alternate routes, might not be a bad idea to have a way to suggest driving slower. I also want to look at adding a minimum speed in some way, since I think driving 55 on a 75mph interstate is not particularly safe.

Finally, the AC/Heater usage is encompassed in the exterior temperature setting. I definitely don't have enough data to verify its accuracy yet, so right now it's identical to Tesla heater / AC performance. In general, a lot of these settings end up doing exactly the same thing as just inputting your efficiency / consumption (mi/kWh) and leaving the other settings at their defaults, just split off into multiple parameters for ease of use.
 

·
Registered
Joined
·
63 Posts
Finally, the AC/Heater usage is encompassed in the exterior temperature setting. I definitely don't have enough data to verify its accuracy yet, so right now it's identical to Tesla heater / AC performance. In general, pretty a lot of these settings end up doing exactly the same thing as just inputting your efficiency / consumption (mi/kWh) and leaving the other settings at their defaults, just split off into multiple parameters for ease of use.
My point is that most people won't have any idea how much turning on the AC, the heaters, the butt warmers, or the steering wheel heaters cost in efficiency, so they won't be able to enter the consumption themselves. I think there's a pretty big difference (10%???) between it being 55 outside and having the heater on vs. 55 outside with having the heater off. If you're able to gather that data from the OBD that would be helpful. If not, so be it.

Besides, since the M3 doesn't have a steering wheel heater, so being able to show how much that hurts its range in cold weather can help us Bolt fans when talking to the Tesla fans. (Just kidding, of course; I love Tesla but this is one item where the Bolt is better than the M3 and Tesla isn't going to be able to fix this one with a software update.)
 

·
Registered
Joined
·
33 Posts
Discussion Starter #13
My point is that most people won't have any idea how much turning on the AC, the heaters, the butt warmers, or the steering wheel heaters cost in efficiency, so they won't be able to enter the consumption themselves. I think there's a pretty big difference (10%???) between it being 55 outside and having the heater on vs. 55 outside with having the heater off. If you're able to gather that data from the OBD that would be helpful. If not, so be it.

Besides, since the M3 doesn't have a steering wheel heater, so being able to show how much that hurts its range in cold weather can help us Bolt fans when talking to the Tesla fans. (Just kidding, of course; I love Tesla but this is one item where the Bolt is better than the M3 and Tesla isn't going to be able to fix this one with a software update.)
That's fair, once I have enough data to build up a trend, I'll have to see what I can do about that.
 

·
Registered
Joined
·
33 Posts
Discussion Starter #14
For anyone interested in contributing data, I've put up instructions for setting up webserver upload in Torque:

https://abetterrouteplanner.com/blog/index.php/2018/08/02/contributing-driving-data-bolt-ev/

If you're already using Telek's PIDs, it should still work assuming you pick the same set of PIDs. I had to make some modifications to his PID list, since Torque tries to be too helpful when setting units, and I was getting odd results in my modelling. If you want to keep using Telek's PIDs, let me know so I can cross-check your data and make sure the units are making it into the database right.

The upside to sending data directly via WebServer upload in Torque is that you don't have to remember to save and email your log every time you drive.

The data points that are most useful to the route planning model are driving data at a consistent speed on flat ground, and charging on an 80kW+ fast charger.

Thanks!
 

·
Registered
Joined
·
334 Posts
Put in my Bolt EV data for a trip from Santa Fe, NM to Socorro, NM. Nothing was accurate. At 65 MPH, it claimed over 10 KM per kwh, and showed no chargers on the entire route, except for one Tesla in Albuquerque. Of no value to me, at this moment. Let us know of the adjustments, including basic charging stations that have already existed for a long time.
 

·
Registered
Joined
·
33 Posts
Discussion Starter #16
Put in my Bolt EV data for a trip from Santa Fe, NM to Socorro, NM. Nothing was accurate. At 65 MPH, it claimed over 10 KM per kwh, and showed no chargers on the entire route, except for one Tesla in Albuquerque. Of no value to me, at this moment. Let us know of the adjustments, including basic charging stations that have already existed for a long time.
That's very strange, when I plot the route myself, I get a more realistic plan:


What settings were you using to generate the trip?
If you're talking about Level 2 "basic charging stations", then those are shown, you just have to zoom into a city-level view. Otherwise the map gets very cluttered when zoomed all the way out.
 

·
Registered
Joined
·
33 Posts
Discussion Starter #17

·
Registered
Joined
·
1,105 Posts

·
Registered
Joined
·
33 Posts
Discussion Starter #19
I'm curious as to how many Bolts are in your sample size.
Only 5, which is pretty small, would love to get more, but it's a marked improvement over the 0 (analytical model) and 1 (only my Bolt) I had before. I also don't control yet for other variables like the LT / Premiere differences, or different types of tires. On my radar is a way to make it easier to set that data collection up, would almost need to be its own standalone app, or Torque Plugin
 

·
Registered
Joined
·
1,865 Posts
Only 5, which is pretty small, would love to get more, but it's a marked improvement over the 0 (analytical model) and 1 (only my Bolt) I had before. I also don't control yet for other variables like the LT / Premiere differences, or different types of tires. On my radar is a way to make it easier to set that data collection up, would almost need to be its own standalone app, or Torque Plugin
Great job, so far! I shared some additional logs (plus a 150 A charging session) earlier. I hope they help. I'll probably be putting about 7,000 miles on in the next 30 days or so (if my plans hold up), so stay tuned! :D
 
1 - 20 of 67 Posts
Top