Still struggling with ABRP choosing suboptimal stops

[irwinr]

I'm planning a trip from Texas to Florida. This is what ABRP suggests:

https://abetterrouteplanner.com/?plan_uuid=e7630eff-26e8-4113-bd4d-a50813334a7f

It quotes a total time of 21 hours 35 minutes. However I noticed one stop is a full HOUR as it has me charging to 94% at one of the stops. Charging to such a high percentage is quite inefficient:

This strikes me as odd, so I zoom in on the map to see if maybe there's just a long stretch with no chargers available...

Interesting... So there's an EA station located conveniently halfway... So why on earth does ABRP have me spending an hour in Lafayette and then passing up this perfectly good charger? Perhaps adding it to the route makes it take longer somehow, so let me add it and see what happens:

https://abetterrouteplanner.com/?plan_uuid=979000ee-fb30-4174-809f-e7d863a8ce72

So manually adding the waypoint decreases total charge time significantly, and takes 5 minutes off the total time for the route... What am I missing? Why is ABRP skipping this charger by default?

[Bo-ABRP]

Thank you for this feedback - the issue has been found and fixed!

Technically, we the planner is an optimizer which optimizes a cost. The cost is usually total trip time, but we modify the cost to weigh in preferences and risks. This means that the total trip time may not always be minimized since other factors do play a role.

In this case, the bug was that we calculated the number of stalls completely wrong, which made Target look like a risky choice. When you add the charger as a waypoint, we ignore that risk. Now corrected!

[Witold Baryluk]

According to information I found, this location in Hammond, LA, has 7 CCS chargers. Shouldn't be too risky (it is unlikely all of them are broken, or all occupied, unless things change later based on real data, which should influence the reroute, but not the main planning).

Anyhow, great you are doing these important fixes, and people are scrutinizing plans for any obvious issues like that.

[irwinr]

On 10/16/2020 at 8:09 AM, Witold Baryluk said:

According to information I found, this location in Hammond, LA, has 7 CCS chargers. Shouldn't be too risky (it is unlikely all of them are broken, or all occupied, unless things change later based on real data, which should influence the reroute, but not the main planning).

It's actually only 4 CCS chargers. There are 7 CCS "plugs" and one Chademo plug. You have to remember that most DC fast charging stations have two plugs per station. At Electrify America stations one of those stations has one CCS and one Chademo. The other stations each have two CCS plugs. However only one plug per station can be used at the same time, and if that one "station" is down or occupied then both plugs on that station are out of service.

Still: even at 4 stations this is not exactly a super risky stop. It's unlikely that all 4 are down. Howver I did just have an experience in Columbus Texas where there were also 4 stations, and 3 of the 4 were not working properly. So it's also not exactly a zero risk stop either.

Either way I'm super happy this got fixed. It was creating some really odd plans for me.

Thanks @Bo (ABRP)!