Level 4 and Level 5 self-driving cars are cars that can drive themselves. Level 4 cars can do this in some places or situations, and Level 5 cars can drive anywhere without a person controlling them.
An e-nose is a device that can 'smell' things like smoke or chemicals. Some self-driving cars use it to notice if something bad, like a crash, has happened.
Term
AI
AI means a computer program that can think and learn like a person. In cars, it helps the car understand what's around it and decide what to do.
Airbags are like big cushions inside the car that pop out quickly if there's a crash to keep people safe. They stop you from hitting the steering wheel or dashboard really hard.
Radar is like a special kind of radio that helps the car see other cars or things around it by sending out signals and listening for them to bounce back.
Natural language processing means teaching computers to understand and talk like people do, so they can listen and respond to what we say.
LIVE
Hi, I'm Dr. Lance Elliott and welcome to my podcast series about self-driving cars.
In this episode I'll be discussing the topic of car crashes and self-driving cars.
If you've become interested in learning more about self-driving cars please see my website
which is www.ai-self-driving-cars.guru for further information.
Okay, let's get started. I bet that most of us would agree that if you've ever been in a car crash
you would certainly know that you were in a car crash. This seems perhaps it's certainly obvious,
but do not let your gut instincts lead you astray. In theory you could have a fender
bender and perhaps be oblivious to it. Those seems to raise questions as to how
a distracted driver you must be to not have felt the impact. All in all it's relatively sensible
to assert that people usually would know that they've been in a car crash. When referring to
car crashes for convenience sake let's agree that a car crash is widely defined, encompassing
outright head-on collisions, minor fender benders, side swipes, rollovers, being rear-ended,
destructive pile-ups, and so on. There are approximately 6 million car crash incidents
each year in the United States. Think of this as 16,000 such incidents per day. How do you know
when you've been in a car crash or collision? Generally it's fair to say that anyone inside
a car that accompanies a substantive car crash is going to likely feel it, hear it, see it,
and possibly even smell it. Here's today's intriguing question. Will an AI-based true
self-driving car be able to discern that it has been in a car crash? And what does this foretell
about the advent of self-driving cars? Let's unpack the matter and see. For Level 4 and
Level 5 true self-driving cars, keep in mind there won't be a human driver involved in the
driving task. All occupants will be passengers. The AI is doing the driving. Realize that the AI
is not a robot sitting in the driver's seat of the car. Though there are various research efforts
underway to create a walking-talking driving robot, that's generally not how we will witness the
emergence of self-driving cars. The AI system is actually under the hood, running on computer
processors that might be hidden inside the body of the vehicle or placed in the trunk. The reason
this is worthwhile to point out is that unlike a human driver that can feel a car crash, there isn't
a robotic body that sits in the vehicle and tries to determine the sensations that arise when a
crash occurs. Though, by the way, there are additional sensors that can be added into a
self-driving vehicle that can make this possible to detect. Nor smells of a car crash possible,
though as a heads-up, some self-driving cars are adding e-nose features into their self-driving
vehicles. In theory, a self-driving car could get sideswiped by another car, for example, and the
AI today might be oblivious to this kind of car collision. Upping the ante its conceivables,
self-driving car might strike a small animal head-on and not have detected the collision. There's
a chance to rolling over something or someone, heaven forbid, laying on the street, and the AI
not detecting that this has happened. Now, to be clear about this, there are ways for the AI to
detect a car crash, but of course that requires that the AI developers crafting the self-driving
vehicle have made sure to include the needed features and functions to be able to do so.
Depending upon what sensors are used on the self-driving car and how the AI has been crafted,
there can be blind spots related to being able to discern a car crash incident that
involves a self-driving car. Some self-driving cars are being devised to extensively figure this
out, while others are less rigorously pursuing this angle. One obvious way for the AI to realize
that car crash has occurred would be via the deployment of airbags. Since those capabilities
are already built into most conventional cars, the AI based self-driving mechanisms can tap into
those same systems and utilize those sensors and its indicator that a car crash has apparently
occurred. The video cameras are another way for this to be detected. They're streaming the street
scene and being examined by the AI driving system to determine what's taking place outside the
vehicle. Overall, the AI would be normally calculating the chances of a car crash, not only as
visual data being collected, but by and large self-driving cars use additional sensors such as
radar, LiDAR, ultrasonic thermal imaging, and so on. Once a somewhat vexing question is whether
the AI should try to warn passengers when a car crash is expected to occur via the natural language
processing capability, the AI could potentially tell the occupants to brace themselves. This might
be useful and allow the passengers to prepare for a bruising, or it might unduly get the passengers
agitated and cause them to take adverse actions that might cause them even greater injury or harm.
Here's another twist related to the role of passengers. Suppose a passenger is watching
the traffic and believes that a car crash involving their self-driving car is about to occur.
It seems likely that a passenger in such a setting might yell at the AI and urge the
driving system to take evasive action. Should the AI take into account what a passenger says
about a potential car crash, we would seem to expect a human driver in some similar situation
to be listening to a passenger. At the same time, we would also expect that that human driver would
make a judgment about whether the passenger was right or wrong in their assessment. For some AI
driving systems, it currently makes no difference what a passenger might say about an impending
car crash. There's no provision for receiving or analyzing any such advice offered by a passenger.
In some senses, this is considered an edge or corner case to be dealt with later on.
The viewpoint right now by many is that the AI as the driver, the sole driver, and no kind of so
called backseat driver, is being sought, encouraged, or included. In terms of hearing a car crash,
the odds are that self-driving cars will have microphones inside the vehicle used to aid in
undertaking the natural language processing aspects. Yes, it's feasible that the AI could
potentially hear a car crash as it occurs. Some self-driving cars are even adding microphones
on the exterior of the car too, allowing for hearing of street sounds, perhaps the sound
of a police car or an ambulance siren. Would a self-driving car realize that a vehicle was being
sideswiped by another vehicle? Assuming that none of the sensors mounted on the side of the car
were damaged, this again is a questionable kind of incident in terms of being detectable. The
interesting point here is that if the AI determines that a car crash or collision of some kind
is perhaps imminent, and yet if there does not seem to be any demonstrable result or outcome,
such as a self-driving car coming to a halt, what should the AI do? You could argue that the AI can
just continue driving along. On the other hand, you would assert that if a car crash or incident
was expected, and yet no apparent damage detected, nonetheless the AI ought to come to a safe stop,
doing so out of an abundance of caution. A human driver would seemingly know when their car has
been involved in a car crash or collision. Also, a human driver could stop the car, get out, walk
around the vehicle, and upon inspection it determine what if any damage has been encountered.
Currently, few self-driving cars are outfitted with cameras that look at the self-driving car
itself. In other words, the cameras are aimed outward, rightfully so, seeking to detect what
is around or coming at the vehicle. Do we need to add cameras that can actually look at the
self-driving car, doing so to try to assess the status of the vehicle? One approach being
considered is to use a drone that would be associated with a self-driving car. The drone
might be used for a variety of purposes, such as bringing an item to the car while the vehicle is
underway, or taking something to someone else from the vehicle. It could also be used to help
look at the car to determine whether there is any damage that has been done to the vehicle.
I hope that you found overall this crash course on self-driving car crashes to be useful.
Well, thanks for listening. Again, I'm Dr. Lance Elliott. I hope that you found today's episode
informative. If you're interested in learning more about self-driving cars, please see my website,
which is www.ai-self-driving-cars.guru for further information.
About this episode
Exploring how AI-driven self-driving cars detect and respond to crashes, this discussion highlights the challenges these vehicles face without human senses. Unlike humans, AI relies on sensors like cameras, radar, and airbags to identify collisions, but some incidents may go unnoticed. The episode also considers the role of passengers in alerting the AI, the potential use of internal and external microphones, and innovative ideas like drones inspecting vehicle damage. It raises important questions about safety protocols and the future development of crash detection in autonomous vehicles.
Dr. Eliot explains how AI self-driving cars can determine that they've been in a car crash. See his Forbes column for further info: https://www.forbes.com/sites/lanceeliot/
