ChatGPT gives driving advice

Toyota Rav4 EV Forum

Help Support Toyota Rav4 EV Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

SolarDaveGreen

Active member
Joined
Sep 12, 2022
Messages
28
Location
Santa Barbara, San Jose
Brought to you by ChatGPT:

David asks:
My electric car has a 45 kWH battery, and I only use the charge between 20 and 80 percent of capacity. My car uses energy at the rate of 5 miles per kWH at 55 MPH and 3.5 miles per kWH at 70 MPH. I charge the car at a rate of 45 kW. For a 600 mile road trip, will I complete the trip faster if I drive at 55 MPH or 70 MPH? Please show your work.

ChatGPT answers:

To solve this problem, we need to calculate the time it would take to complete the 600-mile road trip at both speeds and then compare them.

First, let's calculate the amount of energy we have available for the trip:

80% of 45 kWH = 36 kWH (maximum usable charge)

20% of 45 kWH = 9 kWH (minimum usable charge)

Usable energy = 36 kWH - 9 kWH = 27 kWH

Now, let's calculate the distance we can travel with this energy:

At 5 miles per kWH, we can travel 27 kWH x 5 miles/kWH = 135 miles

At 3.5 miles per kWH, we can travel 27 kWH x 3.5 miles/kWH = 94.5 miles

Since the total distance of the trip is 600 miles, we will need to charge our car once during the trip. We can assume that we will charge the car from 20% to 80% during this stop, which means we will need to charge 27 kWH x 0.6 = 16.2 kWH.

Since we charge the car at a rate of 45 kW, we can calculate the time it would take to charge the car:

Charging time = 16.2 kWH / 45 kW = 0.36 hours = 21.6 minutes

Now, let's calculate the time it would take to complete the trip at both speeds:

At 55 MPH, we can travel 135 miles before needing to charge, which will take 135 miles / 55 MPH = 2.45 hours. Including the charging time, the total time for the trip will be 2.45 + 0.36 = 2.81 hours.

At 70 MPH, we can travel 94.5 miles before needing to charge, which will take 94.5 miles / 70 MPH = 1.35 hours. Including the charging time, the total time for the trip will be 1.35 + 0.36 = 1.71 hours.

Therefore, we will complete the trip faster if we drive at 70 MPH, which will take 1.71 hours compared to 2.81 hours at 55 MPH.

WAIT!

David spots an error in the calculations, and asks: Please adjust the calculations to account for a total trip length of 600 miles, which will require multiple charges.

ChatGPT answers:

To adjust the calculations for multiple charges, we need to calculate the distance we can travel on a single charge, including the charging time. We can assume that we will charge the car from 20% to 80% each time, as before.

Usable energy = 36 kWH - 9 kWH = 27 kWH

Distance at 5 miles per kWH = 27 kWH x 5 miles/kWH = 135 miles

Distance at 3.5 miles per kWH = 27 kWH x 3.5 miles/kWH = 94.5 miles

Charging time = 16.2 kWH / 45 kW = 0.36 hours = 21.6 minutes

Assuming we drive at a constant speed of either 55 MPH or 70 MPH, we can calculate the number of charges we will need and the total time for the trip as follows:

For 55 MPH:

Distance per charge = 135 miles

Number of charges = ceil(600 miles / 135 miles) = 5 charges

Total charging time = 5 charges x 21.6 minutes = 108 minutes = 1.8 hours

Total driving time = 600 miles / 55 MPH = 10.91 hours

Total trip time = Total charging time + Total driving time = 12.71 hours

For 70 MPH:

Distance per charge = 94.5 miles

Number of charges = ceil(600 miles / 94.5 miles) = 7 charges

Total charging time = 7 charges x 21.6 minutes = 151.2 minutes = 2.52 hours

Total driving time = 600 miles / 70 MPH = 8.57 hours

Total trip time = Total charging time + Total driving time = 11.09 hours

Therefore, even with multiple charges, we will complete the trip faster if we drive at 70 MPH, which will take 11.09 hours compared to 12.71 hours at 55 MPH.

WAIT!! WAIT!!

David remembers that there is "charging overhead" and asks for one final calculation:

David asks: Please redo the calculations assuming there is 15 minutes of "charging activity overhead" for each charging stop.

ChatGPT
If we assume an additional 15 minutes of "charging activity overhead" for each charging stop, the total charging time for each stop would be 36 minutes instead of 21.6 minutes (36 minutes = 21.6 minutes + 15 minutes). We can then recalculate the total trip time as follows:

For 55 MPH:

Distance per charge = 135 miles
Number of charges = ceil(600 miles / 135 miles) = 5 charges
Total charging time = 5 charges x 36 minutes = 180 minutes = 3 hours
Total driving time = 600 miles / 55 MPH = 10.91 hours
Total trip time = Total charging time + Total driving time = 13.91 hours

For 70 MPH:

Distance per charge = 94.5 miles
Number of charges = ceil(600 miles / 94.5 miles) = 7 charges
Total charging time = 7 charges x 36 minutes = 252 minutes = 4.2 hours
Total driving time = 600 miles / 70 MPH = 8.57 hours
Total trip time = Total charging time + Total driving time = 12.77 hours

Therefore, with an additional 15 minutes of "charging activity overhead" for each charging stop, we will still complete the trip faster if we drive at 70 MPH, which will take 12.77 hours compared to 13.91 hours at 55 MPH.

END OF CHAT

David adds: While it may seem like driving faster wins, the time needed for leaving the freeway, finding the charger, paying for, starting and stopping the charge, then re-entering the freeway, can sometimes take as long as half of the charging time - or more - depending on charger location and how uncooperative the charger is. Consequently,adding the "charging overhead" closes the time difference between the two speeds a bit.

Personally, I much prefer driving slower. It's less strain on my battery and inverter and it's less stressful for me too. I also only need 5 working chargers to make my trip instead of 7. Also, I'd rather be driving at any speed than charging, and the faster trip has me spending 4.2 hours for the charging instead of 3.

These calculations line up very well with the monthly trip I drive in my RAV4 EV between Santa Barbara, CA and San Jose CA.

If you haven't played with ChatGPT, this is the kind of thing that can make it a lot of fun to use.

What do you think of ChatGPT's calculations? Do you see the remaining error?
 
davewill said:
Personally, I think that https://abetterrouteplanner.com/ does this much better and easier.
Agree. Especially since I got them to add the RAV4 EV as a selectable vehicle type. I have the OVMS updating the ABRP cloud in real time, just like the BT dongles and software do on other vehicles.
 
I hadn't tried https://abetterrouteplanner.com/ but I just did. As an anonymous user, it provides some unreasonable default values. For example it shows charging to 94% in one case. Not going to happen. The chargers I am familiar with have slowed down to 12 kW or less at that level of charge, and more often than not the JdeMO green light is strobing, so I don't ever intentionally charge that high. I don't ever intentionally charge over 80% so that when I am charging, it's always at full speed. Perhaps after logging in and fiddling with the defaults that site would be useful.
 
SolarDaveGreen said:
I hadn't tried https://abetterrouteplanner.com/ but I just did. As an anonymous user, it provides some unreasonable default values. For example it shows charging to 94% in one case. Not going to happen. The chargers I am familiar with have slowed down to 12 kW or less at that level of charge, and more often than not the JdeMO green light is strobing, so I don't ever intentionally charge that high. I don't ever intentionally charge over 80% so that when I am charging, it's always at full speed. Perhaps after logging in and fiddling with the defaults that site would be useful.
Yes, there is a plethora of Advanced options. You can see mine has live data available. The ambient temperature reported is obviously not correct. I have not figured out how to get the real data from the car, nor how to not send zero C to ABRP.

ABRP-RAV-Options.jpg


It also keeps track of charging sessions and driving history.

ABRP-Extended-Charge.jpg
 
SolarDaveGreen said:
I hadn't tried https://abetterrouteplanner.com/ but I just did. As an anonymous user, it provides some unreasonable default values. For example it shows charging to 94% in one case. Not going to happen. The chargers I am familiar with have slowed down to 12 kW or less at that level of charge, and more often than not the JdeMO green light is strobing, so I don't ever intentionally charge that high. I don't ever intentionally charge over 80% so that when I am charging, it's always at full speed. Perhaps after logging in and fiddling with the defaults that site would be useful.

Yes, you can override all of the defaults, including the maximum charge level. That said, it knows that charging slows at higher SOC (although it may not know the exact amount that JDeMO does so), and took that into account before routing you that way.
 
davewill said:
...
Yes, you can override all of the defaults, including the maximum charge level. That said, it knows that charging slows at higher SOC (although it may not know the exact amount that JDeMO does so), and took that into account before routing you that way.
1/2 OFF. For example, RAV4EV HV Battery charging current from Jdemo
jdemo_curr.png
 
davewill said:
SolarDaveGreen said:
I hadn't tried https://abetterrouteplanner.com/ but I just did. As an anonymous user, it provides some unreasonable default values. For example it shows charging to 94% in one case. Not going to happen. The chargers I am familiar with have slowed down to 12 kW or less at that level of charge, and more often than not the JdeMO green light is strobing, so I don't ever intentionally charge that high. I don't ever intentionally charge over 80% so that when I am charging, it's always at full speed. Perhaps after logging in and fiddling with the defaults that site would be useful.

Yes, you can override all of the defaults, including the maximum charge level. That said, it knows that charging slows at higher SOC (although it may not know the exact amount that JDeMO does so), and took that into account before routing you that way.
I have sent them the full charging curve of my car on AC and DC, so it's at least representative of the model, but maybe not your specific car.
 
miimura said:
I have sent them the full charging curve of my car on AC and DC, so it's at least representative of the model, but maybe not your specific car.

Excellent! That means that if it's telling someone to charge to 94% like the example, it really is because that's faster than driving to a different charger at some intermediate location so they can charge to a lower SOC.
 
On my 300 mile trip (one way), I have a few personal rules that have kept me from having to scrounge J1772 chargers partway home:

1. If there are 2 or more DCFCs in a location, try to charge at the first one I encounter so if it's broken or busy I can go forward to the next one and avoid backtracking.

2. If there is only 1 DCFC at a location, make sure when I arrive there that I have enough charge to skip it and go to the next charging destination.

3. Check recent plugshare user reports on every planned charger before I leave and charger by charger to make sure they all appear to be working.

4. If there are more than one DCFCs in a location and they are close together, "rotate" my charges through them across trips and trip legs to "keep the door open" by manifesting CHAdeMO utilization in all of them.

5. If the next charging destination has only 1 DCFC, drive slower (hypermile) to ensure I can skip it. If the next destination has multiple (apparently working) DCFCs, drive faster. (OK, I admit it, I am ALWAYS prepared to skip a charging stop).

6. Assume 100% starting charge on the first leg, and 80% charge on all subsequent legs.

7. OK to drive down to around 16-20% remaining charge on the final leg (ends at my reliable home charger - and it's the end of the trip!)

The Hwy 101 corridor between Santa Barbara and San Jose has some challenging stretches where the DCFC spacing is wide and charger reliability is not 100%, but my "rules" have kept me out of trouble for 15+ trips so far.

The app can handle a few of those with the advanced config but not enough to delegate my charging decisions to it. Need more AI ;-)
 
SolarDaveGreen said:
The app can handle a few of those with the advanced config but not enough to delegate my charging decisions to it. Need more AI ;-)
No, really we just need more chargers. When there are reliably working DC Fast Chargers every 50 miles, you don't have to have so many rules.
 
Back
Top