Tesla Model S

[Ronzuki]

Perspective? I live industrial automation, unfortunately, everyday...it's what I've been doing for 30 years across a vast number of industries. Quote, specify, design, build, program, install, commission, maintain, modify, upgrade...the entire 9-yards. And believe me when I say lately it's all become a scary proposition even w/o factoring in all this cyber BS. Plug that RJ45 cable in to a switch, and ALL bets are off. Forrest Gump comes to mind. NO ONE wants to pay the price to do ANYTHING completely, thoroughly and properly. The stuff we have to spec in to jobs (because of cost alone) is all junk, made and/or programmed wherever, by the absolute cheapest sources possible. The bottom line is all that matters and the lowest one wins the job, period. Doesn't matter if it's a critical system or not. Thus, drastically compounding the epidemic as described in UL's white-paper for 'managing cyber security'. UL essentially fired a warning shot across all of our bows. Doesn't matter, we've all known about these risks for a very long time. Nothing is being done to eliminate them simply because they can't be. Cost prohibitive. All under the guise of "increased productivity". Once again...not. Just because you can, doesn't mean you should....technology and automation is all fine and dandy until it stops working. The frequency at which it stops working, for whatever reason, is becoming far too great. And of course, far too costly. Round-n-round it goes...

Just have a look at some of the threads on this forum alone regarding weird crap that has been popping up as of late with the limited tech in our cars.

The one example of wonderful (critical-system) automation we're all familiar with, currently contributing to local congestion, traffic control signals. Stellar example of, in most places, un-managed automation technology in that it is installed and essentially forgotten about. Which is the single largest mis-guided notion bean-counters, across nearly all industries, have about the automation gimmic. Eliminate people (jobs), pay for it once and that's the end of paying for it. Uh, yeah, I don't think so. Better than 1/2 the traffic signals in my travels do not perform proper flow control. Why? Costs money to stay on top of the automation to make it function properly. Who wants to keep paying never-ending increased taxes? Nobody. Besides, any work involved would go to the lowest bidder, and really, what's the recourse for a gov contractor not getting the job done right anyway? That's right, pay him again, dearly, to have another go at it.

'Good-enough' is a wildly popular mind-set when it comes to automation. Usually means the budget has been exhausted and the project, whatever it is, needs to be wrapped up quickly and closed to minimize any bleeding. We used to jokingly refer to this methodology around here as classic half-assed automation. Well, it's the norm now as opposed to the exception.

BTW, there'd be far fewer unexplained accidents if people would stop staring at their crotches while driving, look ahead, and around them, at what's going on and what at they're doing instead...just saying.

[KuroNekko]

I think industrial manufacturing automation is quite a bit different from automated driving technology. These vehicles not only detect their surroundings with sensors, cameras, etc., while navigating with GPS, assisted GPS, etc., they are making calculations and "decisions" to adapt based on various data input. Hence, some of the biggest names in computer and software programming are behind the technology. Also, these vehicles will soon be communicating with other similarly equipped vehicles, being capable to synchronize speeds and braking. The result will ease traffic gridlock by factoring out a major cause of traffic: human error. Numerous studies have showed that human error in calculating appropriate braking and flow speed are the leading causes of traffic flow problems, let alone accidents.

On the topic of traffic, traffic lights are programmed, not automated, during critical traffic congestion times. They are automated (dynamic control) when traffic is light, yielding to the direction with more traffic. It's when sensors and cameras detect vehicles (on a need basis) that traffic flow is changed. However, in rush hour, it's obvious they are set on timers (fixed time control), calculated by humans, for a controlled flow of traffic. Again, the flaw is on the part of the human when the conditions are at their worst given not all cities utilize coordinated or progression traffic lights. This is very evident when you drive in different cities and notice their different traffic flows.

I have no doubts on your expertise in the field and certainly understand your cautionary tone, but the reality is that human error is the greatest threat in just about anything. From aircraft crashes to car accidents, it's human error or disregard that is the leading cause of accidents, not failures in automation or mechanical components. For example, as a new motorcycle rider, it was greatly and repeatedly emphasized in the MSF course about the greatest risk to a rider: inattentive drivers. It's not the rider, the motorcycle, or road conditions that pose the greatest risk, it's other motorists who aren't paying attention to the rider. Intersections are the locations with the highest accident occurrences (for any vehicle) and there's a reason for that: cross-traffic with other vehicles. Intersections elevate the human decision-making element and hence it's no surprise why so many people get into accidents in intersections. Basically, the more control humans seem to have over a situation or scenario, the more likely they #@*& up. It's because not everyone cares to put in the attention and care, especially when others are more vulnerable like motorcycle riders.

So again, it's a matter of perspective. Some people don't trust technology. Some people don't trust government. Some people simply don't trust others out there. However, the statistics are very clear in indicating the weakest link in just about anything. From aircraft crashes, car accidents, industrial accidents, etc. it's human error that's the prevalent cause. I'm far more concerned about the lack of training and competent people than the greater implementation of technology for automation.

[Ronzuki]

Really? Hacking is hacking. It's not a question of trusting technology. I don't trust the humans involved in any of it any longer. It boils down to reality and fact. What do you think the water treatment systems we've designed control systems for are comprised of? And then, because they can, municipalities will physically connect those systems to network computers with ZERO regard for vulnerability even after being warned there were no safeguards quoted, designed or commissioned in the stand-alone control system. How about Electrical Power Distribution systems? Falls under the very general definition of industrial. How about pharmaceuticals? How do you think those nifty life saving (or ending) drugs come in to existence? What's a little hack of a formulation going to do? Ooops, my bad. I will say of all the industries I've worked in, over-all, they're the most stringent with every aspect of their process controls (because they have the dollars to be that way). However, there are those in that industry as well that aren't. Want some interesting reading, Google Stuxnet.

Another case of just because you can, doesn't mean you should:
http://www.msn.com/en-us/money/technolo ... ar-BBy5YUX

May not be "industrial", and this is my entire point, but do you think the guy that's got one of those in his chest might be a tad concerned that his own personal critical system is vulnerable? Bet that'll keep him up at night. But hey, what the public doesn't know can't hurt them, right?

A snippet from the white paper:

Cyber security and the Critical Infrastructure
A principle focus of efforts to protect against cyber security threats has been on those entities that are considered part of the critical infrastructure. The critical infrastructure can be defined as “IT assets, networks, services and installations that, if disrupted or destroyed, would have a serious impact on the health, security or economic well-being of citizens and the efficient functioning of a country’s government.” Industries most often designated as part of the critical infrastructure include defense, energy generation and distribution, water systems, transportation and shipping, financial services, healthcare and public safety.
Predictably, the central role of critical infrastructure industries in day-to-day life means they represent an attractive target for Cyber attacks. The Industrial Control System Cyber Emergency Response Team (ICS-CERT) of the U.S. Department of Homeland Security estimates that 295 incidents involving cyber attacks against critical infrastructure operations were reported in the U.S. in 2015, a 20 percent increase from the 245 incidents reported in 2014. Cyber attacks reported in connection with manufacturing operations accounted for a third of all reported incidents (97), followed by energy sector entities (46 incidents, or 16 percent) and water systems operations (25 incidents, or 9 percent).
Because of their overall importance, entities that work within designated critical infrastructure industries are expected to comply with mandated or recommended cyber security requirements and practices. In the European Union (EU), for example, the European Programme for Critical Infrastructure Protection (EPCIP) focuses on critical infrastructure entities within the transportation and energy industries and details specific requirements applicable to entities within those industries. These requirements include the development of an operator security plan that identifies important infrastructure assets, provides a detailed threat assessment based on asset vulnerability, and details countermeasures to combat cyber threats. Some EU Member States, including Germany and the United Kingdom, have additional cyber security requirements applicable to critical infrastructure entities.

This is just about what is known and what has been reported. I can guaran-damn-tee you there are faaar more incidents than what have been reported...for obvious reasons. Admission is evidence....because, ya know...they're expected to comply.

First I and many other people I know have zero use for this non-sense in an automobile. Don't want it, don't need it and don't want to pay it. Secondly, I don't trust auto makers, or their suppliers, to do the right thing when it comes to 'cyber-security' of automobiles, period.
VW fiddling with their TDI car's software is proof of that. There's already been documented hacking and take-over of auto ECMs. This is nothing new. So why make it more prevalent?

[SamirD]

This is some great reading as always, and I think both of you are right, with the best path being somewhere in the middle.

I've seen my fair share of non-tested wahoo our there--just look at most software out there, which ironically can also be the software that runs the automated hardware. It's still prone to problems because it has the human element in it, just in a more powerful way. And it's problematic because there's zero effort put in to test any of it. Humans are plenty, we can just be our own guinea pigs, right?

The standards have definitely slipped, but there is evidence that those that uphold higher standards than 'getting by' win the race. The Apples and Teslas of the world have pushed the envelope, but only with enough testing that their products consistently deliver the performance expected--and you can't do that without some careful testing. But that being said, there's always issues because, unfortunately, humans still make mistakes.

It's always the early adopters of bleeding edge technology that get cut (or their head chopped off)--while in the long run, everyone benefits from the vetting of the bugs and the technology becoming mainstream. I think the allure of money is making companies try to fasttrack this process, but there's a real danger to rushing this process as it brings immature technology to the masses too quickly, usually resulting in the loss of money or even life.

[SamirD]

So on a slightly different note, I got to test drive a P100D with ludicrous speed. This was an autopilot v2 car, and they don't have the software working on those yet, so no autopilot. But it was definitely a fast 0-90mph, on par with my Corvette. I even nailed it taking a right turn from a stop sign to see how crazy it would get--not much drama at all--full control because the power was so quickly regulated to each wheel that there was barely time for the tires to even start breaking traction. Simply amazing.

[Ronzuki]

Still not ready for prime-time....



but hey, at least it realized it crashed itself and IMMEDIATELY activated the hazards! I notice any of these damn self-driving cars on the road I'm getting the hell away from them.

[KuroNekko]

Still not ready for prime-time....



but hey, at least it realized it crashed itself and IMMEDIATELY activated the hazards! I notice any of these damn self-driving cars on the road I'm getting the hell away from them.

Tesla's Auto-Pilot wasn't designed to entirely replace driving. In fact, by the definition and standards of other governments (like Germany), it's more akin to advanced cruise-control than actual self-driving vehicle technology. Hence, while it's not very assuring that Auto-Pilot couldn't properly detect the temporary barriers, the real problem was that the driver wasn't driving or even paying attention.

I think the misconception about the Model S is that it's a self-driving car. It's not. It just happens to have one of the best technologies for it, but it's not fool-proof. The car really should be driven by the driver.

That being said, I'd trust a Tesla with Auto-Pilot 100 times over the conventional car being driven by someone looking at their smartphone instead of out the windshield which, by my observations, is about half of those on the road.

[Ronzuki]

Technology not fool-proof? (say it isn't so). Being misused? Mislabeled?
I'm sure it states clearly what you are inferring amongst all of the legal CDW (Caution/Danger/Warning) gibberish contained within the owner's manual that no one has read. Or, if they did, chose to ignore what they read, or, didn't comprehend what they read, or,... If one were to actually read it and comprehend it, they'd have to question why they'd spent so much money on the car in the first place. How was the vehicle marketed to them? Hmmmm. Well then, if this technology isn't intended to be 'driverless', then the technology should simply disengage itself from the Enhanced Cruise Control mode when one's hands are no longer on the tiller for some brief period of time, aye? There's that lack of exception handling not going on in the design and programming of all this nifty stuff. How did the government allow this to happen? Who's to blame? Like I said though, the flashers came on instantly after it wrecked itself (it's all about safety right?). I'd wager to say the only one's to gain in the aftermath of that video, and the many, many more sure to come, will be the lawyers.

My observations are in-line with, if not worse, than yours regarding those 'drivers' constantly staring at their crotches (or wherever) or have a gadget stuck to the side of their head in one fashion or another. Regardless, I trust neither.

Further, to no surprise, I am a proponent of all 'devices' (of a particular technology) being rendered 100% inoperable in a running motor vehicle, moving or not, anywhere on any public road system (the horror, how will we survive?). It'll never happen since the techno-drug is far more addictive and in greater use than any illegal substance out there.

Vehicles sitting at off ramps and not moving when they should be particularly pizz me off. This of course just adds to the dangerous congestion and traffic flow control problems that are already plaguing the area. Especially when my arse is one of the ones still sitting on the main thoroughfare waiting to exit and vehicles, with yet more questionable operators behind the wheel, are whizzing by me at speeds up to 70mph (in a 55mph zone). Now, you'll quickly realize that the off-ramp's traffic flow technology fails to observe that I and many others are sitting there, stopped in a very long line, ON a 4-lane divided highway waiting for the inadequate technology to cycle the light on the 2-lane overpass with barely a handful of vehicles approaching from either direction. Cause and effect. Expensive Technology (mobile device drugs for the crotch gazers) further compounding problems with even more expensive failed, outdated and poorly conceived traffic control technology.

So, that Tesla, and/or its naïve operator, have just caused another unnecessary traffic tie-up as well as inconvenienced a few folks likely with car debris stuck in their tires, all for no good reason. Technology has its uses and its place and I am of the strong opinion that the average Joe's automobile isn't one of those places. Simply can't handle the exceptions to everyday driving.

[KuroNekko]

I think the use of Auto-Pilot in a Tesla should be modeled after the real auto-pilot on aircraft from which it got its name. When a passenger plane takes off or lands, the pilot and co-pilot are in control of the aircraft as these are the most important times during a flight. However, once at cruising altitude and conditions are normal, auto-pilot is engaged to take over. To my understanding, at no point does the pilot and co-pilot get disengaged from the responsibility of flying. They aren't getting drunk or rushing into toilets for renewing memberships to the Mile High Club. They are present in the cockpit and monitoring the aircraft and air space, but simply not physically flying it.
Also, when conditions occur like turbulence, what do the pilots do? They manually take control and often alter the altitude to get out of the turbulence air pocket. The same sort of perspective should be taken with unusual road conditions and elevated-risk scenarios in cars driven by Auto-Pilot or other automated driving tech.
I think this practice behind the real auto-pilot should be the model for Tesla's Auto-Pilot and it's my understanding that it is. However, it's the owners who have other intentions.

Does this mean we can't trust new technology? No. Sorry, but I will hardly ever side with that opinion as I will often see the weakest link as the fault; human error. People often misuse technology and delegate it to do what they are responsible for. Hence, I won't necessary fault Tesla or Auto-Pilot for these kinds of accidents. Had the "driver" been even half-aware, they should have taken control to avoid the temporary road barriers.

I must also say that my background is also why I'm pro-technology. Having grown up in Japan for most of the 80's and 90's, I have firsthand experience at technological advancement done well. I went to school on a man-made island via an unmanned monorail that was never late or missed exactly where it was supposed to line up on the platform. It even operated reliably within weeks of a major earthquake (magnitude 7.3 on Richter scale) after crews thoroughly checked it for damage and compromises to safety. And get this: this was over twenty years ago.

Compare that with today in our nation's capitol where I work. The Metro subway system is so unreliable and unsafe that they literally had to close down segments of lines for parts of the day to implement repairs stemming from a safety and maintenance crisis. Some lines are running one or two trains an hour. This has caused such huge delays that ridership dropped significantly as more people took to driving their own cars instead to commute. This then caused more traffic but that didn't delay the road construction on the 295 freeway, the only freeway going through DC. The traffic and parking congestion is much worse than it used to be.

So is it always technology we should be afraid of? I don't think so. I was raised in a country that had to resort to it for their challenges from a lack of choice. It's often not the technology, but the implementation of it and how it's utilized. Misuse it and it should be no surprise that things can go wrong. However, if implemented properly and carefully, it can truly progress various issues to make them safer, more efficient, and less costly.

That being said, I respect your cautionary tone and think it's a respectable and valid opinion. However, as you also expressed, it's human error that is often the most dangerous threat. You can improve cars, trains, aircraft, and infrastructure, but you can't cure stupid. If technology can mitigate that however, I'm all for it. Tesla or not, that driver would have crashed into that barrier if they weren't paying attention.

[Ronzuki]

We are, more or less, in agreement. You've just supported my point a bit further with the above info, but, don't seem to mind all the garbage and expense that comes along with technology. I like my technology, certainly I'm not "afraid" of it. I frustrate with it everyday. What I am afraid of is the way it gets used. It's simply evolving to a point that's becoming far too costly and dangerous. I don't like it when, and where, use is not really required, needed or justified. Just because you can, doesn't mean you should. But that's always been man's problem.

Japan is Japan, and here is here. The air base my son was stationed at in Japan operated just as though it were plopped right here somewhere in the good ole USA. Oh the stories. He couldn't wait to leave. He's missing Japan, but not the people on and running that base. Japan had major roadways re-built to better standards and levels of quality after the tsunami, in mere days, than any American could ever hope for on a planned and funded road project here. Here, a crappily done roadway 'improvement' milking takes years and usually goes into the exponential over-run territory as far as costs are concerned. NEVER what was budgeted for. Forget nations, might as well be two different planets when it comes to people...and the technology. Pretty much why I prefer a Japanese designed and built vehicle.

People. You say you trust the technology and, not so much, the people. I simply ask, who designs, builds, programs, implements, maintains (not really) and uses all of this technology? R2D2? People. People get too lazy and misuse nearly everything. That's fact. The more of it there is, the more it gets abused...it's a drug. An enabler of sorts.

People. What single factor primarily drives people's decisions around the world? Money. Technology done right costs boat-loads of money that no one wants to spend until ABSOLUTELY necessary. And only after something goes very, very wrong. Cyber security (there's a joke). Takata airbags. Life-saving (and ending) technology. That debacle is so monumental, it'll never get resolved. A problem that was known about looong ago, and yet, the faulty technology was still being installed in to hundreds of thousands of vehicles. Bad technology decisions made by who? People. Why? Money? The older cars that are on those recall lists will be made into toasters and recycling bins long before they ever get around to making enough new "safer" dashboard grenades to "fix" them all in a timely fashion. Have to spend lots and lots of money to fix that technology in a timely fashion, and quite obviously, no one is interested. Technology, People, Money. Let the courts and lawyers sort it all out. Newer cars are still being recalled with those faulty damn things in them, it's criminal.

I trust me driving my vehicles (operating the machine), myself, and not getting in to a situation that requires an explosion in my face far more than I'll ever trust all the technology behind that explosion. Nor do I want to incur the cost of it. If I could turn them off, I would. Side airbags in the Heep were an option at the time which I did not want and happily did not pay for. Based on the shear scope of the recalls, nearly everyone is driving around at one point or another with a frag grenade nestled right there between their hands (if they're not fiddling with some electronic distraction gadget), in their face, and rolling the dice. People are scared of loaded weapons in their house...and yet, they'll happily jump in front of one several times a day. Government mandated technology is the most offensive technology of all.