SoylentNews is people
SoylentNews
Many customers at a small town grocery store in Alberta had trouble unlocking their cars, while others had their car alarms going off mysteriously, since early January. According to the CBC, after trial-and-error of turning off electricity to nearby buildings and calling in the federal government for assistance, the source of the interference has been traced back to "faulty consumer electronic equipment stuck in transmit mode".
The story of the original mystery can be found here.
Key fobs that suddenly won't unlock vehicles. Cars that won't start. Alarms that go off for no reason and can't be quieted. Something mysterious is thwarting drivers outside a grocery store in the small Alberta town of Carstairs — and it's sparking all kinds of theories.
The problems have been happening for weeks in the parking lot outside the Westview Co-op grocery store in Carstairs, a town of about 4,000 about 60 kilometres north of Calgary.
[Ed. note: key fob: A passive wireless electronic device that usually uses RFID technology to control access to buildings, containers, computers, etc. by being placed near a detector.
Remote keyless system (RKS):
Widely used in automobiles, an RKS performs the functions of a standard car key without physical contact. When within a few yards of the car, pressing a button on the remote can lock or unlock the doors, and may perform other functions. A remote keyless system can include both a remote keyless entry system (RKE), which unlocks the doors, and a remote keyless ignition system (RKI), which starts the engine.
Updated: 20190202_131403 UTC]
(Score: 5, Informative) by pipedwho on Saturday February 02 2019, @08:54PM (2 children)
In case anyone is wondering how this can happen: it’s not due to traditional high power occlusion of a weaker signal, but a problem with the way most digital receivers are designed.
What happens is that the receivers ‘hear’ a continuous incoming stream of otherwise valid preambles and headers that they detect and lock on to. So when the (much more powerful due to proximity) actual fob tries to transmit, the receiver sees it as corrupting interference since the receiver had already locked onto the preamble and header of the foreign interferer. So the data of neither packet is received for that event.
However the bad transmitter is continually transmitting making it about 99% likely to be first to be detected, so there is still a small chance that if you hit the button on the fob enough times it will eventually work. This is because the frame can be received once in a while if the correct devices are close enough to overpower the interferance, and the timing is such that the receiver rejects a foreign frame as bad and then restarts the preamble detector just as the correct fob transmits. It locks to the correct fob first and since the interferer is far enough away to be insignificant as a source of additive noise, the correct packet won’t be corrupted.
For ‘proximity’ based keyless entry/ignition, the fob periodically transmits a beacon. However, since a high percentage of the beacons will be lost as per above, the car is likely to determine that the fob is ‘out of range’ and therefore the car will remain locked and/or will not start.
This can be overcome with a properly designed receiver/protocol, but I’m guessing most of these systems are block copy pastes of the chip reference schematics and sample source code implemeted by a guy fresh out of university.
(Score: 3, Interesting) by sgleysti on Sunday February 03 2019, @03:57PM (1 child)
<cough>Time To Market</cough>
At one time I assumed that electronics engineers working in a professional capacity at companies really knew their stuff deeply and understood all the nuances of the things they worked on, like I aspired to do in my hobby projects and in a project I worked on at a startup. Then I landed such a position at a larger company; god, what a revelation. The M.O. seems to be: Quickly absorb just enough information to get it working. Sometimes it's even worse. We had trouble with some optoisolators turning off too slowly, and this younger engineer spent two weeks futzing with resistor values before I explained the basic equations governing optoisolators and where to look in the datasheets of adjacent components to find his voltage threshold targets. He hadn't done any of the basic groundwork.
My manager says I'm really good at technical stuff, which somewhat surprised me. Here's what I bring to the table: I have a pretty good mental model of digital and analog circuits, I read the datasheets, I use circuit theory (and later simulators) when doing design, and I use the scientific method when troubleshooting. This stuff is just the basics, not anything special (or shouldn't be special).
(Score: 2) by pipedwho on Sunday February 03 2019, @11:47PM
Luckily, I've never been employed in a division of a company that made those sorts of demands on me. Although, there were some that might have tried it once or twice. But, in critical system designs, most layers of management will defer to technical expertise, most of time anyway.
However, over the years I've been called in to 'fix' many failures caused by this mentality. The problem is too many companies hire people who don't even know what they don't know. And in a lot of cases, "good enough" turns out to be "not quite good enough".
There are many 'gotchas' in digital comms system design (especially wireless), and it's easy to get something that fails catastrophically if everything doesn't go smoothly. Crypto and cybersecurity is another related area where engineers unfamiliar with the field can be completely unaware of potential 'dangers' of a naive design/implementation.