> “There’s just massive amounts of data being collected, and if you want the service, you have to agree to it. And people do agree to it, and they have relatively little control over things,” Scassa says.
I'm curious what this will do to the used car market in 10-15 years. Say the original owner consented to all this data tracking. They go to sell the car to buy a new one, and the 2nd owner does not agree to the tracking. Is there a vendor-provided way to shut this off? Or is the vehicle always stuck in a "data collecting mode" regardless of who it is sold to?
Furthermore, if a customer changes the car to stop data upload, that could be seen as tampering with a vehicle. This could lead to legal consequences, but I'm not a lawyer so I can't dwell on the specifics.
This is not legal advice, and you should consult an attorney if you believe that an OEM may come after you for tampering with your own vehicle.
That said, I believe that there is very limited legal recourse for a vehicle OEM who wants to prevent an end user from changing the data collection and upload features of their vehicles. The worst thing would be potentially voiding the warranty, but for 10-15 year old used vehicles, that's not an issue.
I work as an in-house counsel at a vehicle OEM (likely not one you've heard of). Our end customers are generally large companies, so it may be different for personal passenger vehicles, but we actively plan for the contingencies where our end users switch out their telematics module, or request to have the telematics data stream be directed to another service provider. Further, in our market, the idea that the customer wouldn't want telematics at all is not really likely, but we also plan for that possibility.
So, I don't believe the used market isn't as big of a deal as you may think. The real concern for personal vehicles is losing essential warranty or service options for new vehicles by turning off the data collection.
Don't forget that Tesla is now in this space, and turning off Tesla's access to it will turn off your car, period. I have a hard time believing that others in the industry will not take the blase reaction to this and then do it themselves.
> Don't forget that Tesla is now in this space, and turning off Tesla's access to it will turn off your car, period.
Which is super scary. A couple years ago one of their firmware engineers had his NDA expire, and he posted some stories about his time there in some forum.
One of the stories that stuck with me is they were really sloppy with their updates. One time the pushed one that wasn't fully cooked (it might have bricked some, I can't remember), so one of the engineers literally wrote a script to SSH into each of the cars to roll back part of the update.
This was really iffy even then. If the car is connected by cellular, it doesn't have a unique IP address to SSH into. If it is connected to WiFi, the router has to be configured to forward that port.
Since the author remains anon, I always thought it was just a troll.
The whole thread doesn't hold water. It was just a mix of plausible, implausible, jargon and some profanity for relatability.
Full quote:
> model s and x use openvpn to talk to their backend. inside that backend there are metadata services that feed info to the system, one of those things being a ~20MB+ (generated by the worst erp system) json payload that describes supercharger shit for the map in the touchscreen. somebody was smart enough to do automated linting but forgot to validate against the custom parser the car runs which caused a segfault in the qt app that runs the ui, which in turn for a variety of reasons forces a reboot of that component. I think we clocked about 15 seconds before it read the file and faulted after boot. it was doing that for an hour before everyone panicked and got me and qa on the phone to fix it. i wrote a quick python/fabric script that ssh’d to as many cars as possible at a time to rm the file
- "On that note, China has a law in place that mandates all electric cars send real time telemetry to their government servers - Model S/X/3, NIO cars and any other electric car if they're driving already complies with that law to be road certified"
- "Don't be surprised if that becomes a mandate in other countries"
Yeah, fair - that's definitely a difference between vocational vehicles and personal vehicles. If we tried to do that, our end customers would laugh us out of the room.
However, I believe consumers also have a way to laugh Tesla out of the room, eventually: public policy ideas like the first sale doctrine, or just straight up passing new laws saying that when you buy a car, the manufacturer must provide basic certain basic services (such as allowing the vehicle to run, and warranty for basic components) even if the purchaser elects to deactivate some of the data collection.
In the past few decades, we have really allowed power to accrue to large corporations, but I think that things like the right-to-repair movement show that the pendulum may swing the other direction, at least in some contexts. Regardless of overall political or economic ideology, as soon as a large portion of the population realizes that these OEMs are basically saying "yes, pay me a good chunk of your annual salary, and I'll let you use this vehicle as long as you don't piss me off," I think things will swing very quickly in favor of consumers. The reason it hasn't yet is because most people don't have to deal with vehicles with these restrictions.
Which is why Tesla will never have a place in my fleet. I can afford it, and love the mechanical tech, but the lack of rights and privacy is appaling. It shocks me people pay for their own piecemeal enslavement for the latest shiny thing.
>That said, I believe that there is very limited legal recourse for a vehicle OEM who wants to prevent an end user from changing the data collection and upload features of their vehicles. The worst thing would be potentially voiding the warranty, but for 10-15 year old used vehicles, that's not an issue.
>I work as an in-house counsel at a vehicle OEM
If you're really a lawyer, I'm surprised you'd write any of this, because it's blatantly false. The Magnusson-Moss Warranty Act of 1975 makes it blatantly illegal to "void" a warranty for any modification, unless it can be proven that the modification caused the issue that the customer is making a warranty claim for.
I wasn't aware of that, but perhaps it will hearten you to know that I don't deal with warranty law in my day-to-day. So, I clearly have more to learn - not sure why you felt it necessary to call my credentials into question over it though.
But in any case, your post does strengthens my main point, which is that fear over surveillance of vehicles by OEMs, with no consumer recourse, is somewhat overblown.
EDIT:
Just read Magnusson-Moss, and I have a question since you seem to know about it - you said the act "makes it blatantly illegal to 'void' a warranty for any modification, unless it can be proven that the modification caused the issue that the customer is making a warranty claim for."
But the act itself says "the warrantor shall not impose any duty other than notification upon any consumer as a condition of securing remedy [...] unless the warrantor has demonstrated [...] that such a duty is reasonable." 15 U.S.C. § 2304(b)(1).
Based on what's actually in the text of the law, I could imagine Tesla or some other OE making the argument that it is "reasonable" to condition the warranty on the consumer continuing to feed the stream of telematics data to the OE, because it allows them to identify warranty issues before they require substantial repairs, thus potentially saving significant money for the OE, and significant time for the consumer - certainly a reasonable thing. I personally don't agree with that argument, but I'm curious how you would respond to it, and how you can be sure that a court would shoot it down.
I only questioned your credentials because you said you were in legal counsel, and at an automotive OEM of some kind. Most people who've done any significant DIY car work know about this law, because it's protected them for decades from dealers that would try to deny them warranty coverage for completely unrelated non-dealer repairs or maintenance. For instance, a dealer can't claim that some kind of engine trouble isn't covered because you replaced your own brake pads, or worse, because you replaced the radio. They also can't deny warranty coverage just because you had an independent mechanic work on your car instead of the dealership. Furthermore, they can't require OEM parts to be used, unless they provide them for free. I would have thought anyone working in automotive legal would know about this off the top of their head.
As for your example, that sounds pretty contrived and ridiculous to me. Doesn't mean someone might not try it though; companies have tried lots of ridiculous legal tactics before, such as Oracle's current API lawsuit. But I imagine any decent court would shoot it down pretty quickly. Warranty claims are made because parts fail before the warranty expires; the absence of telematics isn't going to magically make some mechanical part fail faster. It might help identify it sooner, sure, but it's still a defective part for failing that quickly, so I don't see how the consumer disabling telematics absolves the manufacturer from covering this. In short, it's never been "reasonable" before this for car companies to have telematics to keep their machines working properly through the warranty period, so why is it suddenly required now? It's not reasonable for anything else either; does a new house need telematics for the house warranty to be valid? How about a blender or toaster in the kitchen? I don't see this argument going far at all.
Fair enough. Thanks for the reply, that all makes sense.
I have never run into a situation where my current company tried to limit its warranty in legally interesting ways (and it wouldn't necessarily get to my desk anyway), so I never really had call to dig and find Magnusson-Moss. I suspect that knowledge of that law is probably more top-of-mind for DIYers than for OEM lawyers.
I still think that whether requiring telematics is a "reasonable" condition for a warranty is more complicated than you think, but what you're saying is exactly the counterargument I would make.
And, I hope I've provided some assurance that I am, in fact, a lawyer. Whether I'm a good one is perhaps up for debate. :)
I agree it should be an opt in policy. This is why I'm against tollways because they don't immediately just charge me and toss the data after say 90 days. They keep where you've been essentially forever or until DOT forgets to back it up and the drive crashes.
Seems to me answering those questions will be part of a greater process to balance power and interests between manufacturers and consumers/users.
E.g., Tesla already seems to try and decide the questions in the way manufacturers would likely have them, i.e. suppressing data transmission is tampering and the problem of car reselling doesn't apply as cars cannot be resold anymore without manufacturer involvement.
I think it's up to consumers to offer a different perspective here.
> s tampering and the problem of car reselling doesn't apply as cars cannot be resold anymore without manufacturer involvement.
I suspect this will become a huge societal problem if it continues like this in a car-dependent country like the US. Yes, the middle-class people will probably still be able to purchase $30,000-40,000 cars for the foreseeable future (even adjusted for inflation), but what will happen to the people who won't be able to purchase a second-hand car that costs more than $5000-6000 (or even $1000-2000)? How will they continue to get to work in a country that values public transport so lowly?
Good reliable cars are currently cheaper than ever. Of all the problems stacked against the poor, I'm not sure that cars have become less obtainable. Adjusted for inflation, the cheapest economy cars are quite inexpensive, and generally speaking, they last much longer than they did 20-30 years ago.
That's the opposite sentiment that I've heard recently (from an American perspective). My understanding is that our used cars are more expensive than ever due to:
a) Certified pre-owned and similar programs making some used cars more appealing (and thus more valuable), removing them from the supply of unvetted, cheap used cars.
b) More folks being able to afford used cars around the world. It seems like at the lower end of the market, many American used cars are shipped to Latin America or the Caribbean (in Europe many older cars are sent to Africa), again reducing the supply of used cars.
This is my sketchy recollection of things I've heard and read, I don't have any direct sources to link you to, but I've certainly heard multiple people bemoan that say, $5-10k doesn't buy you nearly as good a car as it used to (even adjusting for inflation).
EDIT: As an aside, the issue of shipping cars to different regions could also prove interesting for owners down the road. There are already some issues there - I believe my car's navigation system is for North America, or maybe US+Canada. So if my car gets shipped to Argentina in 10 years, the Nav will be useless or will need an update. Not a huge deal. But if Tesla's Autopilot is tuned for conditions and laws in a certain region, it would presumably require reprogramming for another region. Would that be an automatic over-the-air update? Or would Tesla not want to bother with free updates to a 20-year-old car?
There's a very long tail of used cars. When you are talking CPO (certified pre-owned), you are talking 0-3 years old, with perhaps 50,000 miles or less (15k-30k probably the sweet spot). Sometimes these can be a good deal, sometimes you can get like 5-10% off the new price (which seems not worth it...to me)
The OP talking about a $5,000 car is talking about a 10-20 year old car with 100,000+ miles.
Both cars appeal to a different set of buyers. The CPO car effectively competes against new cars.
It's not uncommon at all today to see a car tick over 200,000+ miles, so buying a car for $5,000 with 100,000 miles on it can be seen as a pretty good deal. Usually people would recommend a good Japanese / Korean car (Toyota, Honda, etc.).
The market was disrupted a decade back by cash for clunkers, which took lots of cars off the road. That raised prices and forced banks to write loans for older cars.
At the same time, emissions controls changed the market in that cars are more governed by use not age. I mostly buy Hondas and know the market a bit, and they are all good for about 200,000 miles without significant maintenance.
I sold a 2003 Honda Pilot with 250k miles for $2,500 last year -- really high considering all of the stuff that will break on a car of that vintage. In the 90's, that pricepoint would be for a much younger car. A place like CarMax will sell a 2010 Pilot with low mileage for $15k, which works because the TCO works out -- you won't have lots of maintenance.
>a) Certified pre-owned and similar programs making some used cars more appealing (and thus more valuable), removing them from the supply of unvetted, cheap used cars.
That seemed to be true for a while, but I've seen prices relax quite a bit lately. It's possible this was just local, and I'm not seeing a broader trend.
It's also the case that perfectly good brand new cars can be had for $13-$14k, (Nissan Versa, Ford Fiesta) which I appreciate are not attainable for many people.
You would think so, but afaik there's almost no older smart-phones around (think 4-5 years-old) because they're expensive to repair, that is if you can repair them at all.
That's why you don't see less-wealthy people commuting to their jobs in an early 2000s Mercedes S-Klass or in a VW Phaeton (even though they'd be able to purchase such vehicles at today's prices), but instead they use an older Honda or a Toyota or a slightly newer Nissan.
Smartphones aren't cars. Cars are much bigger, therefore easier to make [somewhat] modular. Furthermore, the recent hype with smartphones was smaller, thinner, and more glue.
To judge by what I've seen lately cars have started becoming full of tablets and tablet-like screens, those things will be expensive if not impossible to repair in 10-years' time.
Sourcing a CANBUS module for a 10+yo car is already a roll of the dice. Take all the modules that are required for the vehicle to operate, are matched to the VIN, can't be re-flashed, aren't available aftermarket and aren't made by the OEM anymore and anything that breaks where those sets intersect turns the vehicle into a brick. The list gets a little longer every day and won't get shorter until we get right to repair legislation with teeth.
You can drive a car without all the fancy electronic doodads. Everyone who drives a 20+yo car does it every day. You can't drive a vehicle that won't start because it doesn't see some quasi-essential (e.g. ABS) module talking on the bus.
>Take all the modules that are required for the vehicle to operate, are matched to the VIN, can't be re-flashed, aren't available aftermarket and aren't made by the OEM anymore and anything that breaks where those sets intersect turns the vehicle into a brick.
I'm starting to doubt this more and more. Dumping the ROM from a read-only chip and flashing a new one isn't rocket science.
After watching stuff like http://www.youtube.com/watch?v=yqWhId-tQXs&t=9m31s
and guys like Dave Jones and Louis Rossman, it looks like the entire repair industry is long due for disruption.
Sure, finding old or obsolete parts will still be a problem, but not impossible.
It's already expensive. The Entune head unit in my Corolla would cost me $1000 for an OEM replacement (they have a known touchscreen issue), but an aftermarket head unit with way more features may run me ~$400 including a mounting kit, and that's if I splurge on a nice one. I may give up a few features by switching, but certainly not $600 worth of features.
> E.g., Tesla already seems to try and decide the questions in the way manufacturers would likely have them, i.e. suppressing data transmission is tampering and the problem of car reselling doesn't apply as cars cannot be resold anymore without manufacturer involvement.
Not that I doubt that they believe this, or even that they've said or implied it, but is there any official statement from Tesla to the effect of prohibiting, or even trying to limit, resale without Tesla's involvement?
The trouble is that this isn't true for large companies who are consolidating all the farms anyway. They can afford the the new equipment. This is just a niche thing for the family farm which is steadily dying off as they can't compete on price. Sure there will always be some family farms, but no doubt they will be driven to say "organic small farm raised" for the whole foods and farmer's market type niche of people. This "heating up" is a drop in the ocean.
> Is there a vendor-provided way to shut this off?
In the Tesla Model 3 and Nissan Leaf there are menu settings to disable it. Additionally in the Leaf it prompts you at interval (every 2000 miles?) to notify you about it.
Leaf it prompts you at interval (every 2000 miles?
Or every g-ddamned time you start the car, depending if you have a first-gen or not. The GPRS radio in ours is unhooked, so now the button press is just ceremony so you can see the screen.
But do note the part where the cell radio is unhooked. Whatever the car, there’s a module in there somewhere that probably just needs to be unplugged.
>Furthermore, if a customer changes the car to stop data upload, that could be seen as tampering with a vehicle. This could lead to legal consequences, but I'm not a lawyer so I can't dwell on the specifics.
Where are you getting this crazy idea? If it's your car, you can tamper with it all you want, as long as you don't violate emissions laws and stuff like that. There's no law preventing you from tampering with your car's data collection or transmission, or most other things for that matter. How do you think the entire aftermarket parts industry exists?
I think the used car market will get destroyed. In a few years buying a used car older than a few years will probably be a quaint memory. That is unless manufacturers will get forced to open up their systems by law.
I think the move to electric vehicles will actually collapse the second-hand car market - no-one wants to buy a car with batteries that no longer hold enough charge to complete a journey. We already see this effect with used laptops and smartphones, however I can never see changing your cars battery pack being a end-user activity.
If batteries get commoditized then it could be an easy process to replace the battery. It becomes no different than many other repairs you would typically make to make a used car sellable.
However it will likely require legislation to ensure there's no DRM bullshit in the battery packs.
It becomes an issue because batteries are so much more expensive than other things you can get though. It's not replacing a alternator, it's replacing a $15000 battery "set". Also I'm hoping electric car longevity will prove out the value in this. Electric car drive trains "in principle" should have much longer lives than gas engines.
Swings and roundabouts in the EU: GDPR should clobber all the nonconsensual data collection, but there is also the "eCall" mandate that will require your car to phone in location to the emergency services in the event of a crash.
(What happens when an EU national drives one of the American surveillancemobiles? Sounds like a straightforward GDPR violation to me, although good luck getting that resolved)
Well, in order to implement it the car has to have a "phone" (mobile station) and a GPS. And I think (although not proven) that in order to meet the time requirement it has to keep both of them "warm", ie registered to network with a GPS fix. Which means at the minimum the mobile operators know where the car is at all times. And the car manufacturer has built almost all the hardware for full tracking.
Anecdotally, I can tell you my 6 year old Nissan Leaf asks me every time I start up the car if it can send data back to Nissan. The 2g radio doesn't actually have a network to connect to, so it's kind of a moot point.
I imagine it'll be like any useful service, once you turn off the data collection, it suddenly loses a bunch of useful features
>I'm curious what this will do to the used car market in 10-15 years.
Depends on LEO. If it is easier to deal with trackable cars, then untrackable cars will be hotter than black tinted windows and radar detectors.
It already is, to some extent; why can I not gut my catalytic converter & change some programming so my car says the outputs are fine? Because the state wants me to get rid of my old functional car to support the economy of newer more expensive vehicles.
> why can I not gut my catalytic converter & change some programming so my car says the outputs are fine? Because the state wants me to get rid of my old functional car to support the economy of newer more expensive vehicles.
But the "manufacturing and shipping a new one" part is temporary until the majority of cars are replaced. The reduced emissions from more efficient cars will be permanent.
Yes, retrofitting would be better, but if that's not possible, what else would you do to get converters into cars?
Also, I think GP was talking about removing a converter from a car that already had one. Literally not doing anything would be better for the environment than that.
Not too long ago I owned a motorcycle from the 70s. I realized at one point that it was probably EMP-proof: no fuel injection, no fuel pump, certainly no OBD anything, and a (literal) kickstarter. The most complicated thing in the wiring harness was the key cylinder and the relay for the turn signals and headlights.
There is an urban fantasy series called "The Rivers of London" series after the first book (or "Peter Grant"-series after the main character) and magic interferes with (in the sense of completely fries) more complicated electronics there, so his master (the last fully qualified wizard of Great Britain) also drives an old Jaguar that is all-mechanical and therefore magic-proof. Very entertaining series with lots of references to current pop culture.
Sounds like Battlestar'ing your transportation. A very popular modification scene in the car world.
I have converted modern cars to points and carbs, its not that difficult. No reason you couldnt drive any car with some motorcycle carburetturs clamped to the intake runners. Hardest part is spinning the rotor and remembering how to set your points. Never drive-by-wire cars will need a throttle cable, and a manual transmission is the only electronics-free option. I have heard a simple mechanical relay logic system can control many autos, but have never tried.
That poor car... It was almost a character of the series and it was always getting the worst end of things. I would drive a beetle if I didn't know it would kill me in an accident. One almost killed my dad.
Yes, you still need an ignition source. The OP is confusing the distributor with individual coils found on modern engines.
A distributor is only there to distribute the spark to the cylinders. On modern engines the distributor has been replaced by individual coils (on per cylinder).
All gasoline engines have a spark source. A magneto, a distributor/coil setup, or a transistorized driver/coil setup in the case of fuel injected bikes. Nearly all bikes from 1920-1990 have a distributor. The differince is that it often runs off, or is part, the stator (generator coil), making it appear to not exist. Just follow the plug wire.
That's simply not true. Every motorcycle that I am aware of (except the old ones with a magneto) has spark plug wires that go directly from a coil, sometimes one per cylinder, sometimes one (with two outputs) for every two cylinders, to the spark plug. They do not go through a distributor. The low voltage wires come from either an electronic module or a set of points (typically one set of points for each coil) which are mounted to one of the crankcase covers, not in a distributor.
And while the ignition cam may or may not be near the alternator or generator, it has nothing to do with it functionally.
Interesting, after some research I find that a few very old motorcycles - Indians and Ariels - did have distributors. But these are a tiny minority as compared to all the Harleys and British and Japanese bikes in the world.
Also, I forgot that some old Harleys have the points in a housing sticking up above the crankcase. It's debatable if this should be called a distributor though, since the spark plug wires don't go anywhere near it.
You can move the points just like you'd rotate a distributor. One set of points drives one coil which provides ignition for two cylinders, one which is firing (near the end of its compression stroke) and the other non-firing (near the end of its exhaust stroke). For a four-cylinder engine, two sets of points and two ignition coils are used.
This setup is even used on modern-ish vehicles, where two transistors drive two coils to run four cylinders.
If your car has points-based ignition, you'll find them in your distributor underneath the rotor. They're switch contact points or breaker points or whatever you want to call them, that are driven by a cam on the distributor shaft. They're closed most of the time, allowing current to pass through the primary side of the ignition coil. When they open, the ignition coil field collapses and you get spark out of the secondary side.
Unless your car is very old, chances are that you have a more reliable reluctor-based ignition, where instead of points you have a magnetic pickup that's triggered by a toothed wheel on the distributor shaft. That pickup drives an ignitor (which may also be inside the distributor, or it might be a separate module) which generates the high current input to the ignition coil primary.
You can twist the distributor to change your base timing, plus you have the vacuum mechanism that moves the points or magnetic pickup to adjust timing according to manifold vacuum, and it probably also has sprung weights that swing out as rpm increases, which moves the cam or reluctor wheel to advance timing.
The actual distribution part of the distributor consists of the rotor and terminals on the cap. That's what routes high voltage from the coil out to each spark plug.
On an old motorcycle, you just have the points (one set for each coil for each pair of cylinders as described above) without the distributor cap and rotor part. You can move the points around the cam to adjust base timing, and you might have springs and weights to advance timing as rpm increases. I don't think vacuum advance was nearly as common on bikes as it was on cars.
Hah, yeah. I like my shitty car projects. I'd like to replace my current daily driver, but I just can't get excited about new stuff, so it hasn't happened yet.
Every motorcycle has to have a distributor, unless it's either so new that it has coil-on-plug like modern cars, or it's so old it uses magnetos, or it's a single-cylinder engine.
That's not true. As far back as the sixties at least, both Harleys, Triumphs and Japanese motorcycles had either one coil for each cylinder or one coil (with two plug wires) for every two cylinders. They weren't mounted on the plugs, but up under the gas tank or under the seat. The plug wires go directly from a coil to a spark plug, so there is no distributor.
The purpose of a distributor on a car was to distribute the output of one coil to several spark plugs. When you have a separate coil output for each plug, there's no need to distribute anything.
I had a diesel car that I once drove with no alternator and no battery: started off a jump and drove to the auto parts store. Radio, windows, and lights were dead, but the car drove fine.
Fond memories of my Merceded W124. You could rip the engine out of the car and it would go on and on. No electricity for motor management required whatsoever.
Yeah - I remember when I was a volunteer firefighter - if you parked a fire truck in an atmosphere that was filled with combustible gasses, the engine would not shut off.
The trucks were equipped with some type of an emergency throw should that situation ever arise. We never had to use it in my time there.
I still own a MB with OM616 [1] engine. It does not need an alternator when it runs. Even to turn fuel supply on/off it uses mechanical valve which is on the dashboard. It turns 40 this year but it starts in a split second even in cold winter.
Diesel ignights via compression and does not require external ignition. For this reason, they can 'run away' and rev until they explode if somehow given too much fuel as the result of a failed injector or oil leak from the turbo that feeds the combustion cycle. The yt videos are intense.
> I realized at one point that it was probably EMP-proof
Different kind of bike, but I used to feel that way about my bicycles. However, sometime last year I was parking at the office and realised even that world is on its way out.
The No-True-Scotsman actual cyclists have largely moved to electronic shifting and such, and the utility bikes are quickly moving to e-bikes.
>The No-True-Scotsman actual cyclists have largely moved to electronic shifting and such
Not true at all.
Electronic shifting is a thing, yes: google for "Shimano Di2". However, road bikes with this are usually at least $4k: it's really something reserved for very high-end road bikes. I got a $3k road bike last year and it doesn't have it, nor does anything else in the low-to-mid range. It'll likely be a while before it comes to low-end road bikes (which are generally still "high end" compared to cheap Walmart bikes)
>and the utility bikes are quickly moving to e-bikes.
Again, not really. There's at least an order of magnitude difference in pricing between a regular utility bike and an e-bike. I've been to cities recently where many, many people commute by bike, and only a very small percentage are e-bikes, though they are getting more popular, but still they're a LOT more money than a regular cheap bike.
Roadies are merely one tribe of scots. The Realest, Truest Scots are the grant peterson/jan heine/PBP types, who have (by and large) yet to accept that downtube shifters are replacable with barcons. >:)
I love those motorcycles. Sometime back I wanted to build my own from parts. Started some research only to find out that some states (in the US) make it very difficult to register a custom built m/c. The assembly seemed easier in comparison.
Two cylinder, four stroke. Old Honda UJM with a small engine and fairly low compression ratio, so you could kick it over in one kick without breaking your ankle.
One of my major requirements when I was shopping for my last car was that it not have a cell modem to enable connected car features. It's too difficult for them to monetize your data if they can't get it off your car in near-real time. I also reckon it's a bit more secure.
The car salesmen have no idea about whether the cars they're selling have modems or not, so I found you have to ask about them indirectly in terms of features. I has success with this line of questioning: 1) Does this car have connected car features? 2) If it does, and I decide to get them later, can you enable them from your computer without me having to bring my car in? If the answer is yes to both questions, the car has a modem.
Unfortunately, my wife wanted a car that failed that test. However, my research had indicated that those features seem to be implemented by a module that's separate from the entertainment system. Hopefully I can physically disconnect it without the car nagging about it being missing, but that's a project for another day.
>Hopefully I can physically disconnect it without the car nagging about it being missing
Don't count on it. Physically locating it and disconnecting it might be very difficult. Worse, since the module in networked via CAN, you literally have no idea how the rest of your car will react. It could switch to limp mode, for example.
Better: unscrew the antenna and screw in a matched termination/load instead. This will protect the amplifier inside (HF reflections due to a miss match could fry it with overloading the transistor), and be an even worse antenna compared to an open SMA (or such) connector.
One thing that's interesting to me is, modern cars know (relatively accurately) whether you're speeding or not. They have GPS and a database of speed limits.
Also depending on what's in the car, they may know who was driving it.
How long before a speed safety campaign starts arguing that car manufacturers should be handing over information about people breaking the law in their cars to law enforcement...
The information is still not even remotely accurate enough for this to work without ridiculously high levels of false positives.
An obvious example would be in Germany. The autobahn for example has sections that are unrestricted, but those same sections are 130kmh (or sometimes lower) if it's between certain times or raining or whatever.
Half of the time my car interprets that as just being 130.
Outside of that; there are countless instances of my car thinking I'm in an 80kmh zone when it's actually 100, 120, whatever. I can override that because I know what the actual speed limit is. Would the car be constantly reporting back that I'm doing a bad?
Ultimately I reckon perfect enforcement of this would just stop me bothering. I'll get someone else to drive me around like an Uber and they can take the endless tickets instead.
I guess that depends on country. For example the UK has no public road (that I'm aware of) where a speed of higher than 70mph is allowed, so that would be an easy initial rule.
Where there are dynamic speed limits in place, it would be plausible to report back possibles (car location and speed) which could then be checked against a central database which was more accurate.
I'm not suggesting that such a system is a good idea, but that I can see car safety campaigners looking on it as the next step in their campaigns...
This already exists and it's a massive fucking joke - it's called "black box insurance unit". My sister had one in her car and during one year we've had to ring up the insurance company multiple times to manually correct her driving score, as she was being penalised for "severe speeding". After asking the company to send us the full tracking data it was completely obvious that the system doesn't have enough accuracy - couple times it tagged her as going 50mph(!!!) over the speed limit, because she was driving 70mph on the motorway going over a small 20mph road.....so the system naively queried the speed limit at coordinates XYZ and concluded that she was going nearly 4x the limit. Good thing her insurance wasn't automatically cancelled over it as I've heard it happen to many people in this country(and then you need to go and fight it with the insurance company, like your time is free or something).
Yep, my feeling for the UK is that's one of the reasons they've not started down this line. The level of convictions for speeding on the motorway would jam the courts solid if they tried it and even a small proportion of people contested the offence.
You would like to think saner speed limits could be agreed as car brakes and safety have rather improved since the 70mph limit was imposed...
Perhaps. The thing is that what you really care about is 40 in a 30/20 sort of stuff.
Which I suppose could theoretically be done far more easily.
If the GPS coordinates are within the ring road of the city and no limit higher than your current speed exists there then you're a baddie and we should bust through your ceiling.
Personally, if we're going to use car telemetry I'd be far more in favour of having the system decide whether someone is actually driving dangerously rather than deciding that 55mph on the North Circular at 2am is worth hitting someone's license up over.
In the US this sort of detection is privatized - by insurance companies lowering your rates if you run their app which beans your location, speed, and motion/fitness data to their servers. At least for Allstate you need 50 drives a month to qualify for the discount, and hard braking or speeds over 80 disqualify a drive from being counted for the discount.
>Allstate you need 50 drives a month to qualify for the discount, and hard braking or speeds over 80 disqualify a drive from being counted for the discount.
The really stupid thing about this is that driving 80+ and the occasional hard braking is just normal traffic condition on some roads at some hours. Penalizing people for needing to use those roads at those times and acting like every other normal human around them when doing so defeats the point of insurance.
There are a lot of useful data points in modern cars. For instance, the compression rate of your suspension can indicate how aggressively you're cornering. This sort of thing is already being used in opt-in programs by insurers in the US. They interpret as much data as possible in an attempt to calculate for each customer a sort of driver safety rating that is then used to determine insurance rates.
I predict that drivers would then be required to retrofit a monitoring device in order to be road legal. Oregon already experimented with GPS tracking as an alternative to paying the gas tax.
Even the very proposals for this say that there should be an ovveride in the cabin, just like you can switch the Start/Stop system off. I'm one of those "petrolheads" and it's not a law I worry about in the slightest.
I rented a van last year from a company I don't ordinarily use and had to phone them up about the constant beeping that I couldn't figure the source of.
Little did I realise that was a direct admission I'd been speeding on my way back to the studio!
- ed
(It wasn't even bad speeding (mostly) - I'm in London after all - it kicked in every time I went a mile an hour over whatever limit I was subject to. Very [very] annoying)
I also suspect that in the near future insurance companies will demand higher premiums from drivers who commute on more "dangerous" roads (i.e. roads on which statistically speaking more accidents happen). The companies already have easy-enough access to most of the drivers' GPS data, they certainly do know the GPS locations of most of today's accidents (via Waze-like services), I'm surprised why they haven't already started implementing this.
In fact there are insurers right now, which give a discount of up to 20% if you share your data with them and if you "follow the rules" like not driving so fast, not doing so much acceleration, …
The problem is, if more and more people switching to those insurances it becomes very expansive for people who mind their privacy.
That is already here for motor carriers. New York, Kentucky and Tennessee have been piloting different technologies for a few years for truck and bus inspections.
Cars are required to have certain telemetry capabilities starting in 2020 or 2021 as well. We’ll have automatic inspections and taxation by the end of the decade.
Law enforcement already has fairly pervasive networks of fixed and mobile LPR. You cannot leave most cities without a record of it, and if you attract federal attention, passage through many interstate corridors is captured with LPR and photos of occupants.
IMO, it's more likely that humans will be mostly priced out of driving, via insurance, once the data show that self-driving vehicles are safer. It still isn't settled exactly how insurance will work on a car without a steering wheel. Will the manufacturer insure its entire fleet or will the customer insure their personal vehicle? Or will there be something entirely new, like a government approved industry standard that shifts liability to the public?
Whatever happens, insurers will make it more and more expensive for a human to drive a car once it becomes clear that it's safer for a computer to do the driving.
I wouldn't mind an automated governor that picks up speed limits from RFIDs in the road. If it could make residential roads safe for kids again it would be worth it 10x over. Although automated fines for passing the speed limit is kinda stupid when you can just limit the car in the first place.
At least in my 2019 Subaru, that DB is already badly out of date for my area. Amusingly, it's all on the high side - 80 for a 75 zone, and 45 for a 35 zone. It's also usually late in registering the switches, by a good quarter mile in some cases.
Talk to automotive guys and you will learn that the most current map with this kind of information is in minimum 6 month old. But most of the time this information is 2 to 3 years behind.
>How long before a speed safety campaign starts arguing that car manufacturers should be handing over information about people breaking the law in their cars to law enforcement...
Probably sometime after we have speed limits reasonable enough that the overwhelming majority people do not routinely exceed them and/or the politicians who write the laws stop driving.
That’s take a long time for this to come full circle and for most people to want such a thing (and being willing to pay for it). Meanwhile, what may become available would be the “Apple Car” or something similar that collects data for processing locally in the vehicle and has measures to not send it “to the cloud”. We’d have another one of those “What happens in your car stays in your car” marketing campaigns!
Fine with me. At this point I'd pay regular price plus the expected discount from data collection, just to be sure. Call it a "privacy fee" and I won't even be mad.
IMO the better question is when data becomes too expensive to be worth it.
There are currently no serious financial downsides to scraping, hoarding, selling, or leaking customer data. There is huge financial upside to doing or risking all of those things.
Companies' data policies may be user-hostile, but they're economically rational.
Data becomes "worthless" when the average joe's phone number is worth $0.003 to marketers, but there's a $500 fine for leaking it, allowing access from an internal team for purposes not specified in a granular opt-in form, or for being found to possess it after its deletion was requested. Routine third party audits are mandatory if you want to interact with other serious players in your industry.
For all the ways that banking and healthcare are hellish, they've at least got the semblance of teeth on stuff like this.
We'll know it's working right when companies spend the same effort to avoid hosting your data that they spend on collecting it today.
> A car camera for Affectiva, a company that is developing technology for measuring driver emotions, in Boston, on Oct. 6, 2018.
No. Just no.
Why would anyone allow collecting this kind of information? There is no way it can't (so, won't) be used against you. Everyone gets angry from time to time.
There is also no way this information is 100% accurate at all times. Emotional recognition is inherently probabilistic -- there will be false positives and false negatives, no matter how good the heuristic is. In the end, it'll be you who has to prove that you were in an emotional stable state, and not the algorithm that gets scrutinized.
Starting in 2021 (or is it 2022?) a driver-facing camera will be required to get a 5 star safety rating. The idea is to catch "distracted" or otherwise incapacitated drivers and prevent accidents, but I am sure it will get used for other things...
You could join the data to billboard locations and sell it to advertisers who want to know what kind of emotional impact their highway campaigns are having.
I'm proud to say that my vehicle has zero tracking, is user serviceable, and my purchase of it had zero carbon impact.
It's also 32 years old, gets 14-15mpg, and drives like a truck. It's a 1988 Chevy Suburban. I bought it a few years ago, for $1450 and really haven't put much into it overall. It runs reasonable well (still some things to figure out) and is comfortable. If I had some money I'd fix some of the things wrong with it and update the interior with seats out of a newer model, but overall, it just works.
And that's what I see going UP in value over time. To think I could own a car that could tell some corporate partner where I went for lunch is just unfathomable. I can't afford to own a car that "cool" and I'm glad for it.
Indeed! It's a matter of what I already have. I didn't go that old because o those things, those things are in place because I bought a vehicle that old. I'm just counting my blessings out loud, that's all ;)
I'm glad you're happy with your ownership of that thing, but I'm more glad that most of them are off the road by now as we've made great strides in emissions and crash safety since the 80s, not to mention mileage. These are factors with measurable real-world impact on your fellow citizens, and, I would submit, more important than your paranoid secrecy wrt your dietary habits.
> These are factors with measurable real-world impact on your fellow citizens, and, I would submit, more important than your paranoid secrecy wrt your dietary habits.
1. Why is this one or the other? Can't we have efficient cars that also don't track us?
2. How, in 2020, do people still not understand how serious location data is? In the USA, and likely anywhere that has cars + little public transportation (like rural areas), your car basically displays your life. Where you live, work, eat, meet people, attend events, etc. How long until law enforcement asks for all people who drove near a protest/rally? Or advertisers get that data and start targeting you based on where you drove? Why should a freaking car company (who is 100% guaranteed going to sell your data) have access to your entire location profile? The thought of this alone is frustrating and maddening because of how easily we hand out our data. People have no idea how powerful bulk collection of user data is.
Heh it's not paranoia, it's just a matter of privacy. Then again, my phone already knows, so it's moot. Crash safety has come a long way, no doubt. But mileage really has not.
For all the complexity of a modern Suburban (at over $50K USD!) it gets 14 MPG around town and up to 23 MPG on the highway. Mine gets about 14/17. Since most of my driving is local, it doesn't really matter.
FWIW unless you have a diesel your official city mileage was 11 when the car was new. But as someone who has to live near other people and use roadways myself, I'm more concerned about tailpipe emissions. NOx standards are literally 100x more stringent starting in 2025 than 30 years ago (50x now compared to 1988).
I used to work for a British car manufacturer which is currently in the process of monetizing all the telemetry they get out of their vehicles, some examples:
* Pothole tracking --> sell to local goverments
* Tyre wear --> sell customer data to target adverts for them / promotions
* Hybrid battery charge levels --> sell data to EV charge point companies.
What was even more concerning was that this data is being collected by another company (who supplied the hardware), who had first dibs on the data.
Will the impersonal multinational corporations that control the software in you car be looking out for your best interests?
Will the data collected about you and your behavior be used in ways that never go against your best interests?
Can individual software developers and teams working for these corporations do much about this if they disagree with it?
The answer to these and many other related questions is probably not.
Little by little, we human beings are gradually building a panopticon monitored by machines under the control of corporate organisms whose behavior is beyond the control of most human beings. As an example, think about about all the people at Google who have tried, mostly without success, to prevent it from acting like every other impersonal corporation. There's nothing preventing it.
Yeah, my last car purchase was an exercise in locating a vehicle without any "smart" technologies. Car manufacturers are the last corporate group I'd trust with anything they are not heavily regulated to provide, such as basic safety measures such as the seat belts they fought for years not to provide.
I plan on this being a car I nurse forever, as it is going to be a generation of extremely intrusive data collection, and extremely poor security on these vehicles. I'm staying away for a complete generation, if possible.
I was in a similar situation and ended up with a 2011 Mazda3 hatchback. The plan is to limp this sucker along until hopefully our public transit infrastructure improves enough that I don't have to drive ever again.
They very well might scan even with no networks configured (some Wi-Fi stacks scan in the background by default; it can reduce the the time the user has to wait when selecting an SSID or initiating a connection), but has anyone claimed that they actually connect? I think that would make headlines.
Agreed. Presumably the only thing they'd be able to reliably connect to is things like xfinity hotspots.
The idea that they'd have some kind of generic login for this would be hilariously insane and definitely make headlines.
But the idea of eg samsung brokering a deal with comcast to get conditional access to comcast modems to upload telemetry data in exchange for a modest access fee?
You could even market this feature directly to consumers: If you're in a house that rents a modem from comcast, your samsung smart TV "just works", and nobody has to fiddle with the wifi password.
There's no evidence that this happens today, but I'd be surprised if it hadn't been pitched repeatedly, and was currently under construction in some form.
As long as self-driving a car will be allowed (and I predict this will become quickly impossible the moment self-driven cars will share the space with autonomous vehicles) I will refrain buying any car collecting AND transmitting data without my consent.
Yes that means in ten years or so I will be riding a used shabby old car.
You can get a new car for $15k in the US. "completely rebuilding" a used car of any quality, by any definition, is likely to surpass 15k in labor alone, unless you're doing the work entirely by yourself, at which point it's likely to cost 15k in tools.
Toyota is also doing this now, they began collecting data all 2020 models with TSS 2.0. They started insurance offerings to generate $$ off of this data, toyotaims.com.
As a benefit for paying for their safety services they will harvest everything by default, it’s not even clear if you can opt out of absolutely everything
I'm curious what this will do to the used car market in 10-15 years. Say the original owner consented to all this data tracking. They go to sell the car to buy a new one, and the 2nd owner does not agree to the tracking. Is there a vendor-provided way to shut this off? Or is the vehicle always stuck in a "data collecting mode" regardless of who it is sold to?
Furthermore, if a customer changes the car to stop data upload, that could be seen as tampering with a vehicle. This could lead to legal consequences, but I'm not a lawyer so I can't dwell on the specifics.