> the car could also self-drive itself to a hospital
I think you're onto a good idea, except for the above quote. It would be a much better idea to have the car automatically pull over to the shoulder, stop, engage the flashers, and call 911. Paramedics can keep you alive en route to the hospital with drugs, cardioversion, and CPR. Your car cannot.
Which hospital is nearest? Which can be reached within the self-driving capabilities of the car? Which has Advanced Cardiac Life Support staff and equipment? Which is still open? Which is not already handling another cardiac emergency? Which route has the least traffic?
Paramedics and EMTs can work these questions quickly and they have lights and sirens to get through traffic. Your car will probably google "hospitals" and (slowly) take you to the nearest one, which might be a closed methadone clinic.
In most cities, an ambulance is going to get to your car much quicker than your car would get you to an appropriate hospital. A modern ACLS rescue unit is effectively a heart hospital on wheels. So calling 911 brings the hospital to you, without putting anybody else on the road in danger from your car's not-yet-quite-ready-for-unsupervised-operation autopilot software.
So what it is needed is that the car communicates with 911 and an automated system will tell it to wait for help, go to a destination where help will go to, or go to a help center (hospital or similar) where help will be waiting.
It could even work if you are at home and have an emergency. Call 911 and if they decide you should go to a health center on your own it will send where to go to your car.
Of course at first wait for help should be the default, but when automated driving improves, and the processing of emergencies improves too, an automated response system should be the norm for most of the time.
Agreed. Technology could help. There are significant elements of assessment, planning, and optimization that now are done manually by emergency responders that probably could be automated.
Very location dependent for response times and traffic. Getting to any ER is better than nothing and the device is unlikely to differentiate cardiac arrest from a ton of other issues that might not need that cardiac team.
Just so people understand, in the present day, what you're proposing involves a $5M mission package with $200M worth of airborne combat control systems and related ground support. Oh, and then the "bring dead guy back to life" bit once they get the intercept.
The idea that we're going to start vectoring ambulances to intercept self-driving cars carrying heart attack victims is pretty crazy awesome.
Sending ambulances to cars where a heart attack victim is detected is, however, a feasible use case.
Where I live, in the order of 10 % of road casualties are actually casualties of heart attack or stroke; there is no actual road accident at all except that the car perhaps stops - but yes, the government still counts them as "road casualties" because that validates car and fuel taxes.
Theres no reason with C2C communication coming online that cars will simply move out of the way if you put the car into emergency mode. Obviously we would build anti-abuse features into the system.
Pushed out further, I don't mind if some rich guy wants me to pull over to get out of his way for a 50 dollar micropayment.
I think you're onto a good idea, except for the above quote. It would be a much better idea to have the car automatically pull over to the shoulder, stop, engage the flashers, and call 911. Paramedics can keep you alive en route to the hospital with drugs, cardioversion, and CPR. Your car cannot.