The following applies to the shared AI (temperature) perception for adjustment of doors, windows, hatches, sun blinds, and heating / cooling functions, collectively referred to as "applications" herein.
Currently the AI, temperature-wise, perceives its environment based solely on the environmental temperature (Weather_Temperature). In reality both the inside (Cabinair_Temp) and outside temperature play a part, with different weights depending on the application.
For example, from experience it appears that drivers would adjust their window predominantly based on inside temperature and less based on outside temperature. Still, outside temperature does play a part (as do a gazillion other factors not currently taken into account) – even when boiling hot inside, when freezing cold outside, both the window's opening probability and its opening degree would tend to be lower, compared to a merely chilly exterior. Temperature-oriented door adjustment (eager opening and closing), on the other hand, seems to depend more on outside than on inside temperature.
Because perception seems to vary depending on the application, ideally we would implement a 1:1 mapping between perception and application; or add an intermediate layer for fine-tuning / adapting the generic perception to each application. Currently there is a 1:m mapping for everything besides wipers, sun blinds, and lights.
Things to consider:
What are the implications of OMSI messing with the inside temperature (#29)?
Can there even be a somewhat rational / "best" approach to perceiving per application? Or is it all purely subjective? And ultimately: Would it be worth the trouble?
Things done so far:
AI passenger perception was changed in 1.5.1 to rely on the weighted average of both outside and inside temperature (as opposed to just the former, which used to be the case), with the former contributing 50% more than the latter. That was convenient though because the mapping was 1:1 to begin with.
The following applies to the shared AI (temperature) perception for adjustment of doors, windows, hatches, sun blinds, and heating / cooling functions, collectively referred to as "applications" herein.
Currently the AI, temperature-wise, perceives its environment based solely on the environmental temperature (
Weather_Temperature). In reality both the inside (Cabinair_Temp) and outside temperature play a part, with different weights depending on the application.For example, from experience it appears that drivers would adjust their window predominantly based on inside temperature and less based on outside temperature. Still, outside temperature does play a part (as do a gazillion other factors not currently taken into account) – even when boiling hot inside, when freezing cold outside, both the window's opening probability and its opening degree would tend to be lower, compared to a merely chilly exterior. Temperature-oriented door adjustment (eager opening and closing), on the other hand, seems to depend more on outside than on inside temperature.
Because perception seems to vary depending on the application, ideally we would implement a 1:1 mapping between perception and application; or add an intermediate layer for fine-tuning / adapting the generic perception to each application. Currently there is a 1:m mapping for everything besides wipers, sun blinds, and lights.
Things to consider:
Things done so far: