Self-Driving Cars' Shortcomings Revealed in DMV Reports (mercurynews.com) 181
A demand from the California DMV of eight companies testing self-driving cars has highlighted a number of areas where the technology falls short of being safe to operate with no human backup. From a report: All companies testing autonomous vehicles on the state's public roads must provide annual reports to the DMV about "disengagements" that occur when a human backup driver has to take over from the robotic system. The DMV told eight companies with testing permits to provide clarification about their reports. More than 50 companies have permits to test autonomous vehicles with backup drivers on California roads but not all of them have deployed vehicles.
It turns out that a number of the issues reported are shared across technology from different companies. Some of the problems had to do with the way the cars sense the environment around them. Others had to do with how the vehicles maneuver on the road. And some had to do with what you might expect from systems made up of networked gadgets: hardware and software failures. The disengagement reports themselves identify other problems some self-driving vehicles struggle with, for example heavy pedestrian traffic or poorly marked lanes.
It turns out that a number of the issues reported are shared across technology from different companies. Some of the problems had to do with the way the cars sense the environment around them. Others had to do with how the vehicles maneuver on the road. And some had to do with what you might expect from systems made up of networked gadgets: hardware and software failures. The disengagement reports themselves identify other problems some self-driving vehicles struggle with, for example heavy pedestrian traffic or poorly marked lanes.
Holy shit ... (Score:5, Insightful)
Holy shit, they've just described, you know ... driving.
This shit is what happens pretty much daily, and for which the act of driving requires you to have a high degree of situational awareness.
Just this morning as I was driving into work, some clown turning off a street into the road I was driving on ... he hesitated, then apparently said "fuck it" and went anyway. Unfortunately he didn't seem to be aware enough or intelligent enough to have noticed me. The end result was I had to pretty much do a panic stop behind some idiot who unsafely pulled out into oncoming traffic, and was pretty much suddenly in front of me and driving at half my speed (which was the posted limit).
Why the hell are these companies acting like they have self-driving cars when they clearly can't handle driving in the real world.
This shit is never going to work if it can't handle random, illegal, and stupid behaviour from the humans on the road. And if they think the world is going to replace all cars with autonomous vehicles, then clearly they expect the rest of us to pay for that future.
This is just hubris from the tech industry who are pretending they're closer to solutions that work in the real world than they really are. The fact of the matter is, we're not all going to run out and buy new cars to allow this awesome future as envisioned by corporations to actually ever happen.
Sorry, but all of the stuff in the above quote is pretty much mandatory for driving a car.
Re: (Score:2)
Re: (Score:2)
At legal speeds in residential neighborhoods, a self-driving car that detected the dog promptly would have a very high likelihood of stopping in time. Brands with crappy sensors would hit the dog, but ones with good sensors you're not going really be able to jump out that fast when the vehicle is going that slow. Brakes are too good at stopping at low speed.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
That's fortunate, since it was your wife's fault.
Human drivers routinely avoid accidents that aren't their fault.
Re: (Score:2)
This shit is never going to work if it can't handle random, illegal, and stupid behaviour from the humans on the road.
Yeah, it's too bad they've stopped trying to improve them and are claiming they're ready for the road now. Oh wait, that's exactly the opposite of what's happening, as cited specifically in the article:
As Telenav [representative of others] put it, “Our autonomous system is still being developed and we are working on improvement cycles. At this stage we expect that (the) driver will be taking over the car control from time to time due to the fact that it is new technology.”
It's easy (also stupid and pointless) to say that if they never improve, they'll never be good enough, because none of them claim they're ready today and all are focused entirely on improving performance.
This is just hubris from the tech industry who are pretending they're closer to solutions that work in the real world than they really are.
I'm aware of at least one company for which that seems to be true. As for the more than 50 other companies
Who could have guessed? (Score:2, Insightful)
It is not like anyone knowning anything about AI could have told them it isn't ready for general roads, and only for special roads. It is not like allowing them on special roads first is exactly the plan in Europe where government listens to experts.
Who watches the watchmen (Score:3)
In the US we define experts as those who are paid by industry leaders to say the things we like to hear.
Re: (Score:2)
In Europe, they define experts as those who are paid by government leaders to say the things the leaders like to hear.
Re: (Score:2)
In Europe, they define experts as those who are paid by government leaders to say the things the leaders like to hear.
Possibly, but fortunately we then have whole lot of governments with different idea, leading to a lot of different experts in debating field.
Re: (Score:2)
I'll tell you since you clearly don't understand how the system works.
I'm either bad at satire or you don't understand how it works.
Re: (Score:2)
Not much I can do if people choose to ignore the little "say the things we like to hear." part.
start with freeway point to points (Score:5, Insightful)
Re: (Score:3)
That's exactly how GM does it: If the GPS sees your not on on a divided highway, no lane assist for you.
Re: (Score:2)
Your knot?
You're not?
Something else entirely? Enquiring minds want to know...
Re: (Score:3)
I love to trigger grammer nazis. it good fun.
Re: (Score:2)
its good fun.
FTFY
Re: (Score:2)
I was just trying to figure out what a "not on" was, and why it would be visible to the GPS. Maybe it is a type of jamming device.
If it was a really a "knot on" "on a divided highway," in my State you end up having to register your address on a list for the rest of your life if you get caught doing that.
Re: (Score:2)
There's one edge case in "city" driving where allowing autonomy on normal streets makes sense and still works: bumper-to-bumper stop & go gridlock where nobody is moving faster than 10mph anyway, and having AI pay attention to the traffic is probably a net improvement over the 90% of drivers who are only halfway paying attention and watching cat videos on their phone in their lap ANYWAY.
One big area for improvement... two-way communication between highway networks and vehicles. It's common for an accide
Re: (Score:2)
The next time you drive in the city consider how many of your decisions are predicated on understanding subtleties (some might occasion "stupidities") of human nature
Well, I can't say for certain how many accidents I've avoided through the human perspective. But if I'm tallying the accidents I and those I know have had driving they're almost exclusively caused by glitches in how we drive, like things you almost always pay attention to but then for some reason we had a lapse of concentration or was tired or got distracted or was emotionally on tilt. If not massive errors in judgement passing other cars that almost lead to head on collisions. In fact, I can't really say I
Re: (Score:2)
I don't think self-driving cars will be perfect, but I think they'll be a lot more consistent and always keep reasonable safety margins.
Probably. However, on the flip-side, they will all consistently fail in the same way if they have the same bug (and bug-free software is impossible), unlike humans, whose errors are a lot more randomized. Sure, not all cars will run the same software, but the market in any given country is bound to be dominated by a single-digit number of manufacturers, compared to millions of human drivers (or a lot more). From a systemic perspective, human variation and imperfection is not a bug, it's a feature.
The other
Re: (Score:2)
However, on the flip-side, they will all consistently fail in the same way if they have the same bug
You say this like it's a bad thing. Once that bug is discovered and fixed, it's fixed in all of the cars. Yes, everyone is going to OTA software updates, like it or not.
Unless it's a hardware issue, of course. I suspect that self-driving cars will get regular updates of "don't do this or you'll crash" warnings that nobody will read. Next model year will have that fixed, recalls might have to happen to update the hardware, etc. But we already do that shit for all of the other parts of the car, so there are s
Re: (Score:2)
Most logical place to start is in a city.
Specifically, bus and garbage trucks.
They are:
]1) Low speed
2) Set routes
3) Perfect for early morning/late night shift. Restrict them to the 2 am to 5 am shift.
4) No need to worry about the state complaining as the state would be the people setting up the rules. They can set everything up with a focus on safety rather than performance.
Re: (Score:2)
True, that; the only thing slower than a bus truck is a house truck.
Re: (Score:2)
Uber is already doing it.
https://www.youtube.com/watch?... [youtube.com]
They're using human drivers to move freight within cities and autonomous vehicles to transport them between cities to depots at the out skirts of the city where human drivers pick them up.
Re: (Score:2)
Seldom (if ever) is there the rather obvious suggestion to limit autonomous vehicles to simple point to point 'highway' trips; but that's exactly where and how it should be done for the foreseeable future, if it happens at all. That is, the (literally) lethal mistake is to introduce autonomous vehicles into the complex and chaotic world of city driving.
Absolutely, self-driving cars should at first be limited only to freeways...not only for the very important reasons you mention, but also because freeways open up an easy (and relatively low-cost) avenue for installing autonomous-vehicle-specific infrastructure along them, which would make them much safer.
However, the companies most pushing autonomous vehicle development don't want to do that, and it's clear why. Use on freeways only would just effectively make autonomous driving a fancier version of cruise
Chip off the old block (Score:5, Funny)
Direct link to DMV reports (Score:3, Informative)
https://www.dmv.ca.gov/portal/dmv/detail/vr/autonomous/testing
Just to say (Score:2)
I told you so.
GASP (Score:2)
Sounds like they have the same problems I have (Score:5, Insightful)
There are times that they can't see the faded lines in the road or the stop lights? Well, yeh. I agree. Same here. And I don't consider that safe either. There are many different inherently unsafe intersections near me. The only difference is that the self-driving vehicle can avoid taking the chance by turning it over to me to take the chance. I honestly don't know if it would be better if it kept control and tried it than handing control to me. But, the buck is passed.
The biggest problem that autonomous vehicles have to overcome is that they expose the faults in a very bad system through extensive data collection. Exposing faults in something people depend on is never an easy road. The messenger often becomes the victim.
Like most who have lived more than half a century, I've been in a number of accidents including:
Driving is dangerous. Four out of four of the accidents above were because of driving while impaired in some fashion - a teenager unprepared to drive in rain, driving while tired, blinded by the sun while driving (should have stopped or greatly slowed), and driving while distraught. And there are numerous other incidents that didn't rise to the level that I would term an accident that would be reported by these self-driving vehicle regulations.
Road maintenance is also atrocious in this country and human drivers die because of it every day. Human drivers are also prone to complain that others shouldn't drive while impaired and then make exceptions for their own needs.
Maybe we should start this conversion by just shining a very bright light on reality - require every new car to be equipped with the sensors to record the reality and disclose every bit of it to the DMV - every solid line crossed, every rolling stop, every time the light is red and we're stil
Re: (Score:3)
The only difference is that the self-driving vehicle can avoid taking the chance by turning it over to me to take the chance. I honestly don't know if it would be better if it kept control and tried it than handing control to me. But, the buck is passed.
Herein lies another important problem which is seldom discussed. "Disengagement" only works as an effective safety backstop if you have an experienced driver at the wheel..
In other words, it only makes sense for the first massively deployed generation of autonomous vehicles, where you can be sure that almost everyone behind the wheel has already spent years driving on their own. However, once you have a generation of people growing up used to being driven around rather than driving, and being driven by the
We're 50 years away from this level of AI (Score:2)
Reason says... (Score:2)
Re: (Score:2)
Trees and pedestrians the death of auto-cars? (Score:2)
Trees block traffic lights and GPS signals. Pedestrians aren't always seen by the bot-cars. So unless you live in a desert, once in a while your car is going to try to kill you or a pedestrian.
There is no way all traffic lights are going to be networked to talk to vehicles in the US. I predict "bot-car friendly roads" will be marked as such, and cars will have to shut down autopilot type driving in all other areas.
I vote we give up (Score:2)
Lets start building the vacuum-powered hamster tubes from Futurama.
Re:what's the plan for moral choice? (Score:4, Insightful)
Re: (Score:3)
I'm sorry but autonomous cars should never need a connection to an Active Directory server.
Re:what's the plan for moral choice? (Score:4, Funny)
I'm sorry but autonomous cars should never need a connection to an Active Directory server.
How else will cars know who is and isn't an authorised user? Are you honestly suggesting we use OpenLDAP instead... what's next, Sendmail to replace the CANBUS?
Re: (Score:2)
It would be like the donkey which dies of thirst AND hunger when placed equal distance between food and water.
Moral choice and the free market (Score:5, Insightful)
A self-driving car's software has the priority to minimize litigation of its creating company.
Re: (Score:2)
It is going to run a damned lot of steering scenarios, pick the one with the smallest projected casualty, and save a log of its calculations for the inevitable trial.
Trust me, the car is perfectly capable of seeing options you as a driver cannot, simply because a CPU runs faster than a human brain for mathematical computations.
Re:what's the plan for moral choice? (Score:5, Insightful)
Trust me, the car is perfectly capable of seeing options you as a driver cannot
That's the opposite of what the problem is. The backing up truck in Vegas, the woman cyclist, the concrete barrier... why can they NOT see things that people can clearly see? Don't use the excuse that *sometimes* people don't see them either; because people can be distracted, and these cars aren't supposed to be.
Re: (Score:2)
I'm not familiar with all of the cases you mentioned, but for the one that I am, it's my understanding that the sensors that were used on the self-driving automobile were inferior to LIDAR, which I suspect may have prevented the accident.
But modern sensors, despite being superior to human senses, still have some technological limitations. You can't blame the software for what the hardware doesn't see.... at least we know that hardware exists with better acuity than human senses which, when it used corre
Re:what's the plan for moral choice? (Score:4, Insightful)
It would be one thing if these companies didn't already know their cars have certain flaws. But the fact is, they know they have vast weaknesses and put them out on the road anyway.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
According to some research, automated cars crash less than human drivers. 1.6 compared to 2.5 crashes per million miles for the most severe category of accidents, and 5.6 compared to 14.4 for the least. The problem is those miles are not really comparable to the entire set of miles driven by humans. I would say the answer is nobody really knows how safe they are compared to people.
https://www.vtti.vt.edu/featur... [vt.edu]
Re: (Score:2)
Re: (Score:2)
Until all the cars on the road have LIDAR, and they start interfering with each other. That's going to be a problem with all types of active sensors - LIDAR, RADAR, ultrasonic. etc.
They all work by sending out a signal and measuring the reflections. That works until there is too much other noise because all the cars around you are also shining their lasers on to the same things you're measuring. Or the laser projections from other cars directly hit the image sensor and throw out the signal to noise ratio it
Re: (Score:2)
One laser can intersect another, and they don't interfere with eachother. They can each have their own modulation and so can be detected individually, regardless of other vehicles using the same technology. Current technology can do millions of these per second.
If car is mistaking sunlight for its own LIDAR signal, then the software is at fault, not the hardware.
Re: (Score:2)
Re: (Score:2)
One of the more recent google self-driving car accidents where they ran into a city bus - the car pulled from a right turn only lane into a traffic lane and didn't see a 30 ton city bus traveling normal speeds.
While I admit I do have blind spots - busses usually don't hide in them - especially when performing what is essentially an illegal moving violation (if I'm going to try to pull off something that dumb I'm usually a bit more alert about what I'm going to hit).
Re: (Score:2)
Re: (Score:2)
I'm British so I am not fully cognisant of US driving rules. Why would a car in a right turn only lane have right of way over a bus in a lane to its left? In the UK, if you change lanes, you have to give way to traffic already in that lane.
It seems to me more likely that the self driving car simply did not detect the bus. Either that or it didn't understand the rules of the road.
Re: (Score:2)
Re:what's the plan for moral choice? (Score:5, Funny)
so if I'm a self driving car with no backup operator, do I prioritize the safety of my passengers? if I have to run down 5 people to keep my rider safe, do I do that?
The car should run over the 5 pedestrians, then the litter of puppies across the street and finally drive off of a bridge. Bonus points if the horn sounds like a maniacal laugh or plays Dixie.
what if I have to do the whole run over your mother / a baby / a nun or run over a bunch of assholes? how are they ever going to solve for this, because whatever it chooses will be wrong
Run over my mother, sideswipe the baby so it goes airborne into a dumpster, and back over the 5 assholes. Preferably spinning the tires on their dismembered corpses. The nun is difficult. I'm not sure if it will piss off more people to hit the nun, or avoid her after all of the other carnage.
You're right, some situations are difficult. Oh wait, avoid the nun and run a bus full of orphans off of a cliff. That will probably piss the nun off to. Problem solved.
Re: (Score:3)
I'm curious what your results on the self-driving car quiz [mit.edu] would be. Hey, if you do well enough, they could use you to program the next generation of self-driving algorithms! Probably on the 'criminally insane' setting.
Re: (Score:2)
Too.
Re: (Score:2)
Oh wait, avoid the nun and run a bus full of orphans off of a cliff.
Yep, username checks out.
Re: (Score:2)
Reducing your kinetic energy should be your first priority.
Re: (Score:2)
Preservation of human life should be your first priority; whatever is required for that (which, admittedly, is typically going to involve reduction in kinetic energy, but will also often involve modification of the path of said kinetic energy). This brings up a really interesting possibility for self-driving cars as the technology improves: the ability for the car to take into account its own safety features (crumple zones, airbags, etc) and the safety features of various things it could crash into in the e
Re: (Score:2)
If the cyclist qualifies as a MAMIL* then run him down. You are doing us all a favour.
*Middle aged man in lycra.
Re: (Score:2)
Self-driving motorcycles will operate under a different set of rules.
Re: (Score:2)
What if increasing it would avoid the accident entirely?
A common response from people who think they're a better driver than everyone else. Just drive (ride, whatever) safely, and not at excessive speed. It's not that hard, and you're not in a rush.
Re: (Score:2)
I've accelerated to successfully avoid an accident before, although not all that frequently. If there's an accident waiting to happen in one place, and I'm somewhere else by then, no problem.
Re: (Score:2)
Re: (Score:2)
so if I'm a self driving car with no backup operator, do I prioritize the safety of my passengers? if I have to run down 5 people to keep my rider safe, do I do that? what if I have to do the whole run over your mother / a baby / a nun or run over a bunch of assholes? how are they ever going to solve for this, because whatever it chooses will be wrong
I don't think it counts. I think it's a simple set of tradeoffs, health > property, life > health, occupants > bystanders.
Ie, if given a choice between damaging the car or hurting a bystander it will damage the car.
If given the choice between smoking a pedestrian or a dangerous swerve into the ditch it chooses the ditch.
But if a pedestrian materializes in the middle of a bridge with no guardrails, well goodbye pedestrian.
Of course, that's a choice 1 or 2 generations down the line, right now it's a
Re: (Score:3)
Re: (Score:2)
I'm not sure that's the best strategy. Consider traveling on a highway at 65 mph. The (human driven) car ahead swerves aside to reveal a nearly stationary vehicle or other large object. Both lanes to the sides are occupied. Is it better to hit the vehicle ahead at whatever speed you'll be traveling after max braking, or gently sideswipe one of the cars to the side? Maybe the car to the side can get over into an empty lane or shoulder and it will be a very mild accident.
Re: (Score:2)
Re: (Score:2)
how are they ever going to solve for this, because whatever it chooses will be wrong
Do what humans do. Step on the brake and hope for the best. The goal of self driving cars is not to solve moral question that don't actually apply in the real world anyway.
Re: (Score:2)
so if I'm a self driving car with no backup operator, do I prioritize the safety of my passengers? if I have to run down 5 people to keep my rider safe, do I do that? what if I have to do the whole run over your mother / a baby / a nun or run over a bunch of assholes? how are they ever going to solve for this, because whatever it chooses will be wrong
No need for moral choices (and no capability for it anyway with machines) ... how about just don't hit stuff?
I think that pretty much covers it? ;)
Re:But (Score:5, Insightful)
Humans are unlikely to improve at driving. Software can potentially be improved on. That's not the same as saying today's software is safer than a human though.
Re: (Score:2)
Humans are unlikely to improve at driving. Software can potentially be improved on.
And the halting problem can potentially be solved too.
Tell you what, lets wait until the potential is achieved before assuming that it will be achieved.
Re: (Score:2)
And the halting problem can potentially be solved too.
I believe it has been proven not to be solvable for general cases. As there is a simple example where halting is indeterminate so we know that the general solution doesn't exist.
Tell you what, lets wait until the potential is achieved before assuming that it will be achieved.
You overlooked the hedge word in my statement. Words like potentially, maybe, or might, indicate a lack of commitment by the speaker. In this case it because I feel there is a lack of concrete proof, and not because I'm trying to use weasel words to make vague statements.
Re: (Score:3, Informative)
Humans are unlikely to improve at driving.
Which human drivers, though?
Looking at road death statistics [wikipedia.org] would seem to suggest that German drivers are a lot better than American drivers, while for instance Russian drivers are a lot worse. Now it's not all due to quality of driving (but also quality of roads, law enforcement, car safety standards, etc.) and the quality of driving itself is impacted by many factors, but we can safely say for instance that drunk driving is a huge problem in Russia that contributes a lot to their death rate being highe
Re: (Score:2)
So if a self-driving car can work in America, it'll probably do just fine in Germany. It'll probably struggle in Russia though.
Re: (Score:3)
In their last month of reported testing ( November 2017 ) Waymo only had 1 disengagement, in 30+k miles of driving, because the car couldn't understand the behaviour of another car.
I am pretty sure I've never driven a stretch of more then 2k miles where I didn't encounter at least one instance of 'unexpected' behavior by another driver.
I'd say that - even today - there at least some companies which have autonomous vehicles that are safer then most, if not all, drivers. I think it's not an exageration to thi
Re: (Score:2)
In their last month of reported testing ( November 2017 ) Waymo only had 1 disengagement, in 30+k miles of driving, because the car couldn't understand the behaviour of another car.
The average driver averages 1 accident every 250000km and 500000km, not 1 accident every 30000 miles. This includes the pool of all drivers, by the way: those who are drunk, speeding, texting, phoning, etc while driving.
I'm waiting for SDCs performance to be as good as the average driver.
Re: (Score:2)
A disengagement is not necessarily a would-be-accident.
Re: (Score:2)
A disengagement is the equivalent of my wife warning me about an approaching object before I see it, even though I keep my eyes glued to the task at hand even on a long trip.
Yes, I see the bicycle, thank you. Yes, I see the pedestrian, thank you. Yes, I see that person in front of us driving like an idiot; there's one behind us, too! Eventually we get to, oh yes, thank you, I didn't see that goat on the shoulder. That's the equivalent of a disengagement. But a goat is unlikely to jump in the street.
If you l
Re: (Score:2)
A disengagement is the equivalent of my wife warning me about an approaching object before I see it
No. A disengagement is when the car has to be actively prevented from killing someone.
(What makes you think your characterisation is the correct one? A disengagement is basically the car saying "I dunno").
Re: (Score:2)
It can be a whole lot safer and Elon was on the right track with the Boring company, just not on target. That is an underground system, in tunnels, a controlled environment, where you would catch a vehicle much like a turbolift and it will take you where you want to go. You could even own your own lift and have a parking spot, would not be cheap. So replace, taxis, public and private transport in major metropolitan areas, every current road would have an underground tunnel system, proving that controlled au
Re: (Score:2)
The test suite can be... fun.
Re: (Score:2)
Add a failing test, add code to fix failing test, deploy to all vehicles - fixed. Humans rearly learn in that fashion.
Indeed, humans rarely learn in that fashion...and that's you know, an advantage. Humans operate in a highly non-linear and unpredictable environment, and they developed, after millions of years of evolution, ways to deal with that. One way is a large variety among humans, so that they do not all have the same flaw (nor the same "fix"). It's really hubris to think that software, with its 60 or so years of development, could more robustly handle the world than humans, in situations that humans have evolved fo
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
They just need to kill more people in their test environments.
Re: (Score:2)
After a passenger eats at Chipotle, the vehicle will automatically set course for the nearest hospital, ...
No. They typically only need the nearest toilet.
Re: (Score:2)
My "land departure warning" system goes nuts at sunset, driving west, when there has been tar poured into the cracks in the pavement. The reflections look - to it - just like lane markers.
Re: (Score:2)
You'll need land departure warning systems when the repainted lines lead your self-driving car off a cliff
Re: (Score:2)
Acme makes self driving cars?
Re: (Score:2)
If they did, maybe the Coyote would have won.
Re: (Score:2)
I saw that cartoon; the self-driving line painting truck ended up chasing its tail right off the cliff, and then a bird on drugs flew past in a Tesla and said something smart.
Re: (Score:2)
Facebook just trained their image recognition "AI" with over 3 billion instagram images
They then only scored 85% in a test.
That is "top 1" accuracy. A label of a close up of a car might be labeled "fender" but the AI's first guess is car. Or a picture might have a girl holding a cat, and the guess of the AI is cat, but the picture is labeled girl. Or a picture of a racoon, but the AI guesses cat.
These are usually either not actual "wrong" guesses, or their wrogness is fairly minor.
See this article comparing human performance to computer performance for "Top 5".
http://karpathy.github.io/2014... [github.io]
Re: (Score:2)
Well, the idea many people have of what a self-driving car is, will have to take a fall, but self-driving cars are coming, they will just not be self-driving all the time, but instead either need constant supervision (bad), or only be self-driving on registered road (most likely highways) where the road is up to spec for the computer to navigate it faster and safer than humans.