Autopilot, as used by the aviation industry, still requires an attentive pilot. It is a pilot-aid to reduce workload, it does not obviate the need for a pilot prepared to maintain control of the aircraft or correct for unexpected conditions. Autopilots can and do malfunction, in this video[0] it attempts to make a >1G maneuver unprompted. You'll note the pilot in command in this video is actively scanning for traffic, he is physically positioned to take control of the aircraft, he is paying attention to instrumentation, and is actively participating on frequency. In other words despite having an autopilot: he is still piloting the aircraft.
I don't think the issue here is Tesla misusing the word. The issue is that the common (non-pilot) understanding of the term is wrong. People piloting heavy machinery have an onus to maintain their currency & proficiency, as well as be ready to correct for faults in their instrumentation and pilot aids.
Autopilot is intended to be a tool to reduce pilot workload so you can focus on other aspects of maintaining correct control of the vehicle. (Namely in an aircraft it assists you with aviation, leaving you better able to navigate & communicate.) Instead what we are seeing with these assists in cars is that people are using these pilot aids and then engaging in unrelated distractions.
If you engage the autopilot in a (certified) plane at cruise level and glimpse over a topographical map of the area, you can practically fall asleep, wake up and you'll still be completely safe (until you run out of fuel). This argument/parallelism with aviation is null.
Airplane autopilot: Engage, divert attention for minutes, not die (consistently).
Tesla autopilot: Engage, divert attention for 10 seconds, die (also consistently).
Airplane autopilot!=Tesla autopilot. Name is misused.
Also take a moment to appreciate what you say "It's ok to use this name on cars, because professional pilots know you can't completely rely on plane autopilots. People should know that and if they die it's their fault."
The video you provided shows a very unusual case. Most autopilots on small aircraft are not really running software so such "bugs" are virtually non-existent, and in large aircraft they only use tested-to-death systems and they practically never bug out.
Do most people think that autopilot is used to avoid other planes and obstacles like mountains? I always assumed autopilot was basically like cruise control for planes and just kept it level, at a fixed direction, and constant speed. Same thing as autopilot on a sailboat.
Old autopilots did exactly that, in addition to keeping a predefined climb/sink rate (long ascents/descents are really boring to do manually in small aircraft).
New autopilots in commercial airliners can do many things (follow GPS tracks or radio navigation waypoints, control engine throttle, line up and bring the airplane 50-100ft over the runway etc) but they never do collision avoidance, landings or terrain mapping.
Airliners also have alarms for low altitude (most have radar altimeters) but they prompt the pilot for action, they're not supposed to avoid anything on their own. Apparently Tesla doesn't even do that.
Modern airliners can and sometimes do do automatic landings [1] [2] [3] [4]. It's done when visibility is very poor. If it is not required, most pilots prefer manual landing. It's less work. The autoland system is more complicated to set up, and is more work to monitor in order to take over if something goes wrong. But when the visibility is low enough, many airlines require their pilots to use it.
Autopilot refers to a wide range of capabilities on boats and airplanes. Many of those require an active, attentive opertator for safety.
The car is warning you if you take your hands off the steering wheel. I don't see how anyone but a moron would value their misconceptions about the term "autopilot" over the clear signals that your hands are required to be on the steering wheel.
I wish people would stop bringing this idiotic point about the term "autopilot" and instead talk about how the different designs of autopilot can encourage or discourage attentiveness in operators.
>I don't think the issue here is Tesla misusing the word. The issue is that the common (non-pilot) understanding of the term is wrong
In fact, because it's well-known that there's a common misunderstanding of the term (as you've acknowledged) but the company chooses to use it anyway, then that's sufficient to represent intentional misuse. They are leveraging this misunderstanding in their branding, then hiding behind the "real" meaning when it's convenient.
Companies test and invest heavily in their branding, which includes a full reasoning of the connotations associated with the words they choose. There is literally no way that Tesla is unaware of people's common misunderstanding.
So, maybe it's clearer if you look at it another way: why choose a word that could create any confusion when there are countless other choices?
>the pilot in command in this video is actively scanning for traffic, he is physically positioned to take control of the aircraft, he is paying attention to instrumentation, and is actively participating on frequency. In other words despite having an autopilot: he is still piloting the aircraft.
If a driver took a similar monitoring posture there are n-situations in which he/she would not have time to react to avoid an accident. There is generally far-greater margin of error and time for correction when an aircraft's autopilot fails. This is why a system that requires such monitoring in an automobile is a fundamentally flawed design. There are too many situations in which there is simply not enough time.
Because drivers are expected to a.) allow the system control of the vehicle but b.) recognize its failures and take back control to correct within milliseconds? That's super-human and, at best, adds n units to the human's reaction time--with potentially devastating consequences.
And, remember, it's beyond "environment monitoring". Drivers must now correct for when the vehicle does not recognize a hazardous situation and also respond when the vehicle itself suddenly creates a hazardous situation (like veering towards a barrier). There is no amount of "environment scanning" that can predict such a malfunction.
Pilots may realize that autopilot is a term that can refer to extremely simplistic systems that require constant pilot attention. The public at large that Tesla is selling to, however, equates the term with big airliner autopilots that could land the plane if needed.
Which is a tiny tiny percentage of the general population.
As has been said it is what the general population understands "autopilot" to mean which is the important part. The fact that the niche and general understandings are different should not be a gap that a company inserts itself into to mislead the public as to their products capabilities.
Yes, most people don't know the nuances of what "Autopilot" means in the aviation industry, but trying to educate the general public is a losing proposition vs just telling Tesla to rename the system.
The FTC should require them to stop using the term Autopilot. The common understanding of that term makes it misleading.
Words matter and Tesla is trying to have the benefit of the term without the responsibility for what it implies.