Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Saving Joshua Brown

This site may earn commission on affiliate links.
Hi.
I am a control systems engineer. Joshua Brown's death was tragic. Here is a quick solution for preventing further accidents like this. I hope it gets some traction.
There is a fantastic product opportunity here. I wouldn't be surprised if someone has either already done it or has it in the works. It would be a natural extension to the product line of Tom-Tom or Garman or Apple or Samsung. That is: a simple device that you throw into your glovebox pretty much like a mobile phone that calculates your vehicle's current GPS coordinates, speed and heading and transmits it on a wireless net to the surrounding vehicles. A simple calculation would then be able to predict a collision. The cheap model would just sound an alarm. The more expensive model would be integrated into the vehicles braking system. This would have saved the life of Joshua Brown. If that truck was carrying such a device even if it wasn't connected to the truck's brakes the Tesla vehicle would have been able to "see" that truck and apply its brakes. If this warning device cost less than 100 dollars it would be an affordable and really cool thing to have in your car to make sure all these driverless vehicles avoided you. You could even throw in a chat function (supported by a cell phone app) so the kids in the back seat could talk to the kids in the surrounding cars. The mind boggles at the possibilities. You could even have an IFF-like squawk transmitting your philosophy of life: "MGB fast mover - death or glory - outa-my-way I waste nerds" ... This would immediately cause your Tesla control system two double the following distance.

I just bought some Netgear to extend my wireless net. It extends the two and five gigahertz bands, has four ethernet LAN ports and one USB 3.0 port. It cost AUD124.00. Most insurance companies would probably be overjoyed to discount you this much on your car insurance policy if your vehicle carries one of these devices. Alternatively Elon Musk should manufacture something like this and give it away. That might silence all his critics.

Does anyone on this list know him? How can we patch this idea through?

It may even be possible to integrate such a function into a mass-market cellphone as an extension of the Wi-Fi hotspot/Bluetooth functionality. Though I suspect wireless net range may be a problem.

There seems to be confusion about what to do with these AI fuelled applications, from a safety perspective that is. There are solutions though. Back in the seventies I was working on systems fitted with what we called "abort programming". The current tech speech is "the kill switch"
(
the thing Elon Musk wants to invent for out-of-control AI - remember the famous line from 2001 a Space Odyssey?
Dave Bowman: Open the pod bay doors, HAL.
HAL: I'm sorry, Dave. I'm afraid I can't do that.
)
Well, the standard solution is: as a completely separate exercise, you task a team to figure out unsafe states of the object under control and develop a simple device that can disable ("kill") the complex automation system if it starts doing dangerous things. The key is that this device must be simple and therefore less prone to failure. This is a proven architectural solution so we DO know what to do. In fact the device I have described above could perform this function. It would also use diverse technology which is another very well established solution.
I could see the bumper stickers now, "Relax. My car is kill-switch-safe"
There is also a role for government here. Government could legislate to make carrying a kill switch compulsory.

Cheers
Les
 
Good concept, certainly as more autonomous cars hit the roads enhancing their performance/accuracy will become more relevant to the masses. Allowing vehicles to "see" each other in multiple ways (radar, cameras and wireless signal) should improve the situation assuming the system is smart enough to sort it all out. Of course there is the problem of system failure. How often does your phone, lap top, Tesla have to be rebooted when it stops working properly?
 
I doubt such a system will be accurate enough to locate you exactly. And just a bit off you get alarms all the time because there was a car parked somewhere or is wasting to get on the highway etc. and then people turn it off because it's to annoying.
 
  • Like
Reactions: Matias
Really? Is that from the NTSB or just pure speculation?

All those facts have been widely published in various news media articles. The DVD was one of the Harry Potter movies. Could all of it be wrong and made up? Possibly, but I suspect it's all true.

And it was 8 speeding tickets in the last 6 years which is not totally crazy. Leaving out the period in which he got those tickets makes it sound like he just got those tickets and it isn't true.
 
Waze is probably the closest thing to this right now that takes in lots of car data to measure traffic. You would really need a compelling reason for people to have these things or a government mandate to get the critical mass required to prevent crashes.
 
Sounds like an automotive version of aviation's TCAS - and with the same weakness: it relies on every vehicle having a working transmitter on onboard. This might be especially problematic when certain portions of the driving public realize both their habits and their speed can be easily monitored with such a system.

Certainly vehicle to vehicle communications will become important as more automation happens - especially the addition of swarm logic allowing a group of cars to drive together, drafting at closer distances and benefitting from the extended sensor reach.

However, in this case I'm thinking it's more important to find ways for the AP car to correctly identify threat/risks, because you can't make certain all obstacles will have transmitters.
 
All those facts have been widely published in various news media articles. The DVD was one of the Harry Potter movies. Could all of it be wrong and made up? Possibly, but I suspect it's all true.

And it was 8 speeding tickets in the last 6 years which is not totally crazy. Leaving out the period in which he got those tickets makes it sound like he just got those tickets and it isn't true.
Understood and my response to him was admittedly abrupt. What I should have said is like most things what you hear at the beginning is usually much different than we learn after a proper investigation. So yes some of that might be true but it's speculative until it's proven by the investigator. BTW we are way OT. :)
 
Like Saghost said, this is essentially the TCAS system used in aircraft applied to automobiles. Great idea, and we're almost there, but unfortunately, I don't think GPS is accurate enough to make it useful in cars. In aircraft, +/- 30 ft is plenty accurate enough because aircraft are supposed to maintain at least 500 ft vertical separation and a lot more than that for horizontal separation.

But automobiles are usually within 10 feet of each other or less in normal operation. For this to really work, you need a locating mechanism that would be accurate to 2 feet or less, and it would have to be adjusted for each specific vehicle because it would need to know the outline/footprint of the vehicle. E.g. in the Florida accident case, the unit would have to know the trailer dimensions and where the unit was located within the 18-wheeler to accurately be able to determine collision probability.
 
  • Like
Reactions: Matias
I don't disagree with the overall concept. I think its fundamentally necessary for mixing autonomous cars along with regular cars, bikes, etc. But, it's something that has to be an industry wide implementation.

What I personally think needs to be improved is AEB systems. They suck a lot worse than what people realize. In this case it didn't activate at all. I'm not even sure how many AEB systems on the market today would have detected this. Maybe the Subaru system.
 
I'd favour a "radar reflector" of some sort - something that makes a "vehicle" more "visible" to the sensors. Hopefully dirt-cheap, and thus could be mandated for installation at the vehicle's next roadworth-test (over here that is an annual test once a car reaches 3 year's old).
 
Hmm...I don't think it would be the right approach..as SomeJoe7777 said one thing is accuracy...just think technically on a curvy road each vehicle crossing you would trigger an alarm, and if someone is crossing you but decide in the last minute to turn left and is crossing your way, there is nothing you could do ... Guess the approach that you have a control unit in your car that gets a good map of what is going around you and if there is always a way where you phisically can go would be a good approach... and that is as far I understand the way Tesla wants to go ...
 
I'd favour a "radar reflector" of some sort - something that makes a "vehicle" more "visible" to the sensors. Hopefully dirt-cheap, and thus could be mandated for installation at the vehicle's next roadworth-test (over here that is an annual test once a car reaches 3 year's old).

Those are allready in use in European trucks; side collision protection bars.

They would have helped JB by giving better radar return from underside of the truck and also JB's car would have hit them with front crumble zone.

Uudelleenohjausilmoitus
 
Last edited:
Brown was driving at a high rate of speed, playing a movie on his DVD player, had 8 previous speeding tickets, and was known as a dare devil with no fear ( Navy Seal). Don't think there was much that could be done to prevent this.
A system that slammed on the brakes when presented with the side of a truck might have done something to prevent it.
 
  • Like
Reactions: Matias
TCAS equipped aircraft are basically self-contained units, no GPS info is necessary. Over the last thirty years TCAS on board aircraft have undoubtedly prevented several air disasters. The issues I foresee translating this technology to automobiles involve reaction time, congestion, and cost. TCAS can analyze and predict future altitudes and trajectories, giving pilots ample time to make mutually effective avoidance maneuvers. TCAS would have a much more difficult time operating efficiently in the split second world involving cars passing in opposite directions. Drivers would likely become numb to numerous false warnings. Like that slight curve in the road. Ditto for heavy traffic. Even for pilots operating in terminal airspace, Traffic Alerts generated by TCAS can be quite pesky. A TCAS type device might eventually provide useful data for vehicles, possibly in a fully autonomous driving world. Most here would probably agree, we won't see that happen anytime soon.
 
Decent idea, but GPS for ground-based vehicles is decades behind in accuracy, especially for all the tiny fluctuations each vehicle does whilst in the lane. Not to mention the security premise of such a system, if hackers ever got a hold of that wireless signal and were able to control the cars braking/control systems.