With all the noise about hacker’s ability to commandeer cars, a though occurs:
Could old-fashioned copper window screen (very fine copper wire forming a grid of ~1mm) be laminated to a car and prevent remote access?
For a specific: Tesla’s OTA firmware updates. And any number of scare-mongering stories about hackers who are into cars over-riding the gas/brake/steering/sunroof/whatever.
The OTA updates to firmware is especially scary: if it can receive firmware updates from Tesla, how long before a hacker figures out how to spoof Tesla’s key?
Anyway, for this thread: Would copper screening form an effective Faraday cage?
A proper Faraday cage is a metal box with no gaps in it. You can use a screen mesh as long as the holes in the mesh are smaller than the wavelength you are trying to block. For wifi type signals, the mesh on your microwave is a pretty good size, so your 1mm x 1mm is probably fairly decent.
Your idea of a copper mesh on the windows has a few problems.
First of all, you need your copper mesh to be in solid contact with the car’s metal body all around the window. Just laminating a copper mesh onto the window is going to leave non-conductive gaps around the outer edge of the window where radio energy can leak in and out.
The car’s body has to be completely metal. Any plastic, carbon fiber, etc. type panels will let radio waves in and out.
You need to seal all holes going in and out of the car. This includes the small gap around the doors.
You have problems with things like the steering wheel shaft, wires, accelerator and brake lines, and other things that go through the firewall or other parts of the car’s body…
All of this assumes that the wifi receive antenna is inside the passenger compartment. If not,then all of this is for nothing.
It would be much cheaper and easier to just disable the wifi receiver built into the car.
Why would a plastic panel be a problem as long as it is covered in the mesh?
he problem with plastic airplanes was a lightning strike - it would not dissipate the energy as do metal planes.
The solution was to implant metal in the layups of the panels.
Why would this not work with auto? The plastic panels will need to be bonded to the screen, as would the metal in the glass (if they can build safety glass with (whatever is between the layers), it should be possible to sandwich a grid in as well.
Or, we could just stop making simply everything “Connected”.
But it will require massive “Oh Shit!” moments before we figure out it is not a good idea to have such vulnerabilities.
There is a middle ground between leaving a car wide open to hackers and encasing it in a Faraday cage (which will also prevent passengers from using cell phones).
I work in the cell phone industry. In some of our labs we have Faraday cage rooms for testing things without interference. The door seals are some what easily damaged and if there is a short gap in the seal around the door even with most of the seal working you will let in noticeable amounts of radio energy. In short I don’t think that you will be able to reliably keep the Faraday cage working without substantial ongoing effort.
Cars can be remotely hacked because they were intentionally designed to talk to the outside world. If your goal is to break that functionality, it would seem easier to find all the receivers/antennas and disable them, instead of modifying the whole car.
It’s not a problem, as long as the mesh is connected to the metal body along all edges. On a car, this would require you to grind off paint around the panel and attach the mesh to the bare metal. Not very good for the longevity of the car.
I ran into this situation just recently at work - we needed a metal box for a prototype electronics box, to provide EMI shielding. But someone made an error and ordered a painted steel box, instead of a bare aluminum box. The lid did not make an electrical contact with the box, and therefore did not provide any electrical shielding. We were in a hurry, so we used a grinder to remove the paint around the edges of the box & lid.
Since we are using cameras and screens for rear-view and side-view, the problem of "visibily = ability to see through a panel is growing moot.
And yes, if we want to use the existing metal as a cage, it will be bonded. Creating the cage would most logically occur at original manufacturing.
Instead of “Don’t worry: Only WE can updated your firmware*, so don’t concern yourself about updates - we’ll take care of everything”, we have “If you want the next update, plug in the auxiliary antenna and enter your code 42** between 05/01/18 and 06/30/18. When update is complete, you will receive notice to remove antenna”.
this update will improve regen braking, not use the pedestrian identification routine to target “accelerate and run over” objects.
** - your code 42 is in your dead tree documentation. When you enter it, there is a code hashed between your car and our transmitter. That hash is in the datastream sent. Your car will not accept any update without both our key plus that hash sequence.
(yes, I know there are better ways to control updates, but mailing out USB sticks might not work any better)
If you have really thin wires, it shouldn’t present much of a visual obstruction. Upthread a mesh size of 1mm was suggested, which would be the gap between the wires; if the wires themselves are only, say 50 microns (0.050 mm) in diameter, then the wires will barely be visible, and you’ll only be blocking about 10% of the visible light hitting the window.
NO. Its not too late. Why should future cars be affected because some few early examples have the problem ?
Tesla should install a “write protect” hardware, eg a switch, into ALL the cars , so that the console ( for navigation, web browsing, sending diagnostics, INFORMING of service requirements, etc, ) cannot actually interfere with the OPERATION of the car. You’d think they’d have that there to reduce their own warranty costs, as if there was no other reason. Is it a white hat hacker activity to go around disabling computers to ensure that a black hat doesn’t get it ?