How does autopilot get so much better efficiency

[Chomlee]

So I just had my MYP now for 3 days. Been driving around and being the engineer that I am, I have been constantly looking at the battery usage and what makes the efficiency go up or down.

I was a little frustrated wondering why I couldn’t find some thing to turn on autopilot until the “duh” moment when I just clicked the cruise control twice. Immediately the efficiency went way up. It’s strange because I am not a harsh driver and was really trying to gently travel to get the best efficiency. Is there some magic in the autopilot?

 

[jcanoe]

Short trips are inefficient as the Tesla vehicle must cool or heat the passenger cabin every time you drive. One way to compare Autopilot efficiency with your driving would be to drive a set highway route (driving both directions) of at least 15 or 20 miles in each direction. Drive this once with Autopilot and once without using Autopilot while driving at the same speed. You can use the Trips function to keep track of the trip mileage and efficiency. (Reset Trip A after you record the results. Many Tesla owners rename Trip B to be Lifetime so they remember not to reset this odometer.)

 

[Chomlee]

Short trips are inefficient as the Tesla vehicle must cool or heat the passenger cabin every time you drive. One way to compare Autopilot efficiency with your driving would be to drive a set highway route (driving both directions) of at least 15 or 20 miles in each direction. Drive this once with Autopilot and once without using Autopilot while driving at the same speed. You can use the Trips function to keep track of the trip mileage and efficiency. (Reset Trip A after you record the results. Many Tesla owners rename Trip B to be Lifetime so they remember not to reset this odometer.)

But is there a known efficiency improvement or is it just me imagining?

 

[Rocky_H]

Is there some magic in the autopilot?

Heh, no. It's just that computer automated tuning is a lot faster and more precise at detection and making adjustments than humans are. It's not about short trips, as @jcanoe was suggesting. It's just driving dynamics.

I am not a harsh driver

You're not intentionally so, but you are a human body with a human foot and ankle, etc. You're going to be overpowering or over-regenerating many times per minute, possibly even a few times per second. And every pair of interactions like that, where you spend too much energy and have to recover it has a little bit more loss.

But is there a known efficiency improvement or is it just me imagining?

No, not imagining. It's probably significant enough to be measurable.

 

[jcanoe]

But is there a known efficiency improvement or is it just me imagining?

You did not provide any data, i.e. Wh/mile results, to compare your driving efficiency without using the Autopilot and with Autopilot turned on. Some Tesla owners may be able, with experience, to drive as efficiently or more efficiently than Autopilot. For example, Autopilot accelerates fairly aggressively from a full stop unless you have selected the driving option Chill instead of Standard (or Sport in the Performance Model). OTOH, IMHO three days is not enough time to master driving the Tesla Model Y for best efficiency.

My post regarding short trips not being efficient, while true, may not have been relevant to your inquiry regarding Autopilot efficiency. Local driving is not a good indicator because of traffic, traffic signals and stop signs. Unless you have purchased Full Self Driving you would have to repeatedly activate and deactivate Autopilot at stop signs and traffic signals. A trip of perhaps 20 or 30 minutes (each way) that includes highway driving where you are able to activate Autopilot for much of the trip route would be a better method of comparing efficiencies. Driving the route in both directions will average any efficiency gains or losses for elevation changes and wind.

 

[Chomlee]

Well I should be able to get some good data in the future. My drive to work involves back roads driving to the swamp with 15 miles of 55 mph on a flat road and no stop signs and no traffic. Of course there will be some difference between morning and afternoon driving (higher ac).

 

[ZenRockGarden]

You're not intentionally so, but you are a human body with a human foot and ankle, etc. You're going to be overpowering or over-regenerating many times per minute, possibly even a few times per second. And every pair of interactions like that, where you spend too much energy and have to recover it has a little bit more loss.

I would not call AP's use of throttle and braking subtle or nuanced when in traffic - it's pretty abrupt at times, and usually smoothness is most efficient...

 

[qdeathstar]

So I just had my MYP now for 3 days. Been driving around and being the engineer that I am, I have been constantly looking at the battery usage and what makes the efficiency go up or down.

I was a little frustrated wondering why I couldn’t find some thing to turn on autopilot until the “duh” moment when I just clicked the cruise control twice. Immediately the efficiency went way up. It’s strange because I am not a harsh driver and was really trying to gently travel to get the best efficiency. Is there some magic in the autopilot?

You probably aren’t as good a driver as you think..

 

[Calliope]

Using cruise control has always been more efficient than not using it -- but it was harder to measure in an ICE vehicle. As Rocky_H said, human responses are not as well-timed or as smooth as a computer's response because: Human. Imperfection comes with the job description.