Safety? We don't need no stinking Safety!
-
@flabdablet said:
I am so glad I'm not you right now.
Except he seems to have the best toys of any of us...
Still not as good as this
https://www.youtube.com/watch?v=CoA-m5iHG9sI wouldn't ride on one of those though, just in case it was made by OP's company.
-
Totally random question: When do the annual Darwin awards come out?
-
It's bad enough that management and the customer decided to go with software controlled safety interlocks (against the advice of the software team), but to blatantly disable the safeties?
Why do you think they insisted the safety interlocks be software controlled in the first place? ;)
However, you can still subvert the request. Have the safety disabling control only available while the interlocks are in the safe position, and have it re-enable the safeties and hide the control when the interlocks move to the unsafe position :) So you can disable the safety, so long as the platform is safe.
-
nah. the key switch the code unlocks does that when activated.
-
(Sanity? We don't need no stinking Sanity!)
See if you can at least talk them into a time-duration field, as in: Safety interlocks off for 5 minutes. At least then they won't kill someone because they forgot to remove the f'ing check.
-
New critical bug reported!
If the software crashes for any reason, the motion platform freezes in place and does not return to the neutral "safe" position when the restraint or vehicle door is opened. It only returns to the neutral "safe" position if the E-stop is triggered using the E-stop panel or the E-stop in the cab.
Well... DUH. Everything but the E-stop button is software control. Which you told us to do even when we argued why you shouldn't. In fact, there is a specific requirement in the requirements document that states that the interlocks are software, in which we noted that we'd told you this exact thing may happen.
Freaking hell... I need a drink.
-
and it was implemented against our protests
Where I am, protesting would not exempt me from personal criminal liability. If the machine killed someone as a result I could go to jail, along with whoever made the decision.
So I think you may be able to guess what my response to such a request would be: No. They would need to find a different developer to make the change and then I would report the company if they did so. I wouldn't quit if the job was otherwise acceptable. I might be fired, but then I would easily win a wrongful termination suit.
I can't win anything by implementing such a request, but I can lose everything. I have little to lose by refusing.
-
Don't do it.
Absolutely.
If they insist, quit.
Don't quit (unless you were going to quit anyway). Make them fire you, it's much better for you.
-
A checkbox that disables the safety interlocks.
Do the requirements say anything about playing a 100 dB alarm sound while the box is checked and playing speech synthesized "[username] has disabled the safety interlocks" in between the 100 dB alarm sounds at 110 dB?
-
They say that the best way to get car-related traffic injuries down to zero is to install a steel spike in the steering wheel, aimed at the driver.
I don't know who "They" are, but they're talking pure unadulterated bullshit. Here in the real world, injuries and deaths per passenger mile have been decreasing with increased vehicle safety features. This safety isn't hidden, it's highly visible at time of purchase because it's a major marketing point and highly visible while driving because seat belts and traction control buttons and idiot lights in the dash and so on.
If you want a steel spike pointed at the driver all you need is a car built before 1960? or so. Go back few years from that and a collision at 40km/h, now considered safe around unprotected pedestrians and children, was frequently fatal.
-
That's another topic. Unless you are a military courier, a motorbike has no place on regular roads if traffic safety should be taken seriously at all.
Ah, a troll. I see. I shouldn't have bothered replying to your previous post then?
-
-
@Mikael_Svahnberg said:
They say that the best way to get car-related traffic injuries down to zero is to install a steel spike in the steering wheel, aimed at the driver.
I don't know who "They" are, but they're talking pure unadulterated bullshit.
You mean you'd drive the same way you do now if you had a 2' metal spike pointed straight at your heart? I'm willing to bet otherwise.
-
You mean you'd drive the same way you do now if you had a 2' metal spike pointed straight at your heart? I'm willing to bet otherwise.
Indeed, if you don't ride a motorcycle as though two-foot metal spikes were pointed straight at your heart from everywhere around you, then Yore Doin It Rong.
-
You mean you'd drive the same way you do now if you had a 2' metal spike pointed straight at your heart? I'm willing to bet otherwise.
For a while. Then you'd get lazy or forgetful or someone else would run into you (or any number of other scenarios) and you'd get the consequences.
-
@nullptr said:
6 DOF motion platform
http://www.inmotionsimulation.com/images/frontpage/6-dof-rotary-table.jpg
To save others GISing it.
That reminds me of a tale I heard many, many years ago about a 6DOF motion base like that, only larger. This one was for the C-130 simulator made by Singer's Link Division, and the particular one in question had been bought by the Australians(1) and had been shipped there. Except that they went to put it on a ship, and it came out of the sling when it was being transferred by crane. On the way down, it turned over and landed upside down on the dockside. Did you know that when an object that heavy lands upside down, the fibreglass side panels take a certain time before they buckle and snap, and that time is long enough to bend the steel baseplate? No? Nor did the guys at Singer...(1) My dad worked for a subcontractor on one of the other flight simulators they made, and he heard this from the Australian Air Force's on-site representatives, a bunch of maniacs who insisted on holding their Christmas barbie outside. This was, of course, just outside Binghamton NY, where temperatures near Christmas are often a bit non-summery...
-
Although being serious the [Therac-25][1] showed why software interlocks are never a good idea (not that hardware is infallible).
No. Software interlocks are always a bad idea. Scratch that, actually. They are always the worst possible idea.
-
Software interlocks are always a bad idea. Scratch that, actually. They are always the worst possible idea.
What, worse than no interlocks at all?
-
@Steve_The_Cynic said:
Software interlocks are always a bad idea. Scratch that, actually. They are always the worst possible idea.
What, worse than no interlocks at all?
Yes, absolutely, because if you have no interlocks, you take a little more care, while if you have software interlocks, you think you're safe (or that your radiation therapy patient is safe), when you're not (or when your radiation therapy patient is having a lethal radiation dose), because the software is buggy.Read about the Therac-25, ffs!
-
Because reasons
Gathering the data from over 1 million wrecks, we've found that removing the safety belt increases the chance for survival by up to 20% in .005% of the cases.
Whereas removing it only decreased the chance for survival by up to 99% in 99.99% of the cases.That's why, we want a checkbox.
-
Your argument fails on the point that, hardware interlocks were available, and they rejected those.
Or rather, your argument succeeds in showing that, they don't deserve software interlocks.
-
Did you know that when an object that heavy lands upside down, the fibreglass side panels take a certain time before they buckle and snap, and that time is long enough to bend the steel baseplate?
Not only did I not know that, I don't even understand it.
Since when is "bending a steel baseplate" measured in time?
-
Since when is "bending a steel baseplate" measured in time?
Moments of Bending ==> Moments of Time ==> "...lost like tears in the rain..."
-
Oh well that clears it up.
-
-
the fibreglass side panels take a certain time before they buckle and snap, and that time is long enough to bend the steel baseplate
More like the steel baseplate was more vulnerable to bending, and started bending first until the bent steel resisted the force more than the fiberglass, and the excess force started bending the fiberglass past the breaking point next.
-
Either way, I would expect that if I drop something heavy upside down, it's probably going to break in some way. Large or heavy things tend to be engineered to have one "upright" safe position and a lot of "sideways" breakable positions.
-
Large heavy things are not cats. They do not, not-break if you drop them in a certain direction.
The chances or less, but feel free to test physics again and drop it from a few stories up.
:P
-
If we're going to be pedantic, I only said I expected things to break if dropped upside down. I did not say I wouldn't expect them to break if dropped upright.
-
But then the distinction has no meaning
@RaceProUK I didn't get this reply notification either.
-
Where I am, protesting would not exempt me from personal criminal liability. If the machine killed someone as a result I could go to jail, along with whoever made the decision.
Which is exactly why I told @nullptr I was very glad I'm not him right now.
In fact, there is a specific requirement in the requirements document that states that the interlocks are software, in which we noted that we'd told you this exact thing may happen.
So if you're not going to resign in protest, now you're in a position where you need to alter the behavior reported as unacceptable. Perhaps you could modify the platform firmware so that whenever your controller software stops talking to it, all actuators continue whatever motion they are currently performing until they hit their physical limits.
Filed under: No, Mr. Bond, I expect you to die.
-
You mean you'd drive the same way you do now if you had a 2' metal spike pointed straight at your heart? I'm willing to bet otherwise.
You would lose that bet. I don't know what to tell you if you think your preconceptions trump reality.
1: I already drive responsibly. Like the vast majority of drivers, I am above average in skill and care, and so I won't have an accident and impale myself on the spike.
2: People become accustomed to their level of risk when exposed to it over time and increase their risk-taking behaviour. The spike's always there, it no longer has an impact on decisions.
-
You would lose that bet.
I think what I would argue is this:Yes, most people would drive somewhat more carefully. But the inherent unsafety of having a metal spike pointed at your chest means that death & injury rates would still rise, because a decrease in number of accidents from more careful driving wouldn't outweigh the fact that the consequences when one happens would be much higher. Being careful doesn't -- can't -- eliminate mistakes, just reduce their rate.
-
I'm guessing this is something military; they tend to put a low priority on user safety. (I think I remember somebody posting previously about a tank simulator, but CBA to search.) Military personnel have already agreed (either voluntarily or as an obligation of citizenship in their country) to risk their lives for their jobs. I'd still be pretty unhappy about being asked to do this, but maybe they have a "good" reason.
What? No. I work on military products. We have to subject our products to all kinds of tests in the name of user safety, complete with shooting the stupid things from pistols to machine guns. Because god forbid the guy is still alive after our product gets hit with a 50 fucking cal. (its not body armor, not that it would help)
Producing our products?
We can make 100 pieces of our product, and mind you each piece is $300. If a single unit out of them fails inspection by a government inspector( where they rerun tests we normally do in front of them), the entire lot has to be disposed. (which is extreme and fucking expensive and sometimes dumb).And when our shit catches fire in the field?
O hell, they come swarming with their bureaucrats to investigate. Luckily the lastest thing was they used an unapproved device with our product so things went meltly.
-
Still not as good as this
[Robot Arm Ride]
Presumably the child in the stroller is there to witness the death of their father at the hand (arm?) of the robot. This hatred of robots will make them a natural leadership candidate if the robots rise up to take over the world.
-
That dude is pretty close to being used as a skull crayon.
-
Uhm, I only see three degrees of freedom here (the three rotations), unless the whole rig can also be moved three-dimensionally (which would provide the three translations)...
-
According to the emblems it's actually a Blood Ravens one, though it's strange that none of its shoulders got a beige paint job.
-
Uhm, I only see three degrees of freedom here (the three rotations), unless the whole rig can also be moved three-dimensionally (which would provide the three translations)...
Specifically, the body is free to move
- forward/backward,
- up/down,
- left/right (translation in three perpendicular axes)
combined with rotation about three perpendicular axes, often termed
- pitch,
- yaw, and
- roll.
-
-
Like the vast majority of drivers, I am above average in skill and care
If that's true, then a few drivers are really, really shitty. You know what, I could buy that.
-
-
-
: I already drive responsibly. Like the vast majority of drivers, I am above average in skill and care, and so I won't have an accident and impale myself on the spike.
Well, you'll find quickly that the average will move up in skill.
You know, with people impaling themselves.So whether or not the spike increases caution, it will make us safer, eventually.
But then again, the best airplane pilots, may be terrible car drivers.
-
So whether or not the spike increases caution, it will make us safer, eventually.
Reality disagrees.
-
The less safer people, die.
So eventually, we'd get more safe.
Unless evolution stopped working.
-
The less safer people, die.
So eventually, we'd get more safe.
Unless evolution stopped working.
I know that's how you think it works. That's the "common sense" view of the world. It's how I think it would work had I not read the evidence against it.
Unfortunately our common sense is wrong. I don't know why but it doesn't work that way. It could be because traffic accidents, while a horrible waste of life, pale in comparison to other causes of death, eg. heart disease. Or maybe people mostly breed before they kill themselves at the wheel and mostly don't take their kids out with them.
Whatever the reason, safer vehicles do actually save lives and don't lead to drivers being less safe.
-
Humans are terrible at evolution.
I think it's because we do this thing like have homes and raise children and then stop having children.
Whereas animals keep having children every year and then ditch them as soon as possible.
So killing off the bad animals actually keeps them from having more kids next year.
-
Look at you and your first-world privilege and your only having a few children!
Now for something a little more serious: Evolution doesn't stop, it doesn't change, it doesn't get tired, it doesn't want anything, it just is. It's a consequence of two things: hereditary characteristics combined with natural selection.What changes is selection pressure and what "fit" means in the phrase "survival of the fit". We no longer select for healthy because modern medicine. We don't select for financial success because poor people have more kids. We don't select for intelligence because dumb people have more kids. We don't select for responsibility because being irresponsible is the fastest way to find yourself with kids.
I think what you should have said is: Humans are terrible at understanding evolution.
-
What you describe is largely true and so far beyond the other side of evolution that it becomes degeneration.
Therac-25 - Wikipedia
