Chalmers: Uh.. a quantum computer? At this time of year, at this time of day, in this part of the country, localized entirely within your kitchen?
Skinner: Yes!
Chalmers: May I see it?
Skinner: No.
Hard to tell... there are a couple of similar (kinda noname) articles on the web, all 3-5 days old. I would have expected this to appear by now on the register, or on ars technica, both followed this story reasonably close...
I definitely don't want to belittle the achievement, but I would imagine that creating safe self-driving helicopters is much easier than creating safe self-driving cars. (At least as long as the skies are just as empty as they are today. Once it's full of things like the roads, that will be a different topic)
I get what you are thinking. Detection, maneuvering and trajectory planning are all much easier than on a road. If you mandate built in transponders collision avoidance is also easy.
But what you are forgetting is everything that isn't normal operations. What do you do if anything fails? A car can just stop, break failures, even steering failures can all be reasonably mitigated. This is not the case when you are in the air. Any failure mode needs fast and accurate reactions, even when critical systems have failed. That is why a passenger plane has two pilots. A modern passenger plane can do most of the flying by itself, yet the pilots need to be there.
Aerospace standards are higher and more difficult to adhere to, ensuring any kind of reasonable safety is extremely difficult. How many of these flying into sky scrapers are acceptable?
This thing has 8 motors. It’s easy to imagine that it could have 8 independent batteries and could be able to perform a controlled descent if two fail. Similarly, it could have redundant sensors and use commodity “majority rules” logic circuits where all calculations are run three times. On top of that, they can already cut over to ground control if the computer can’t proceed.
>This thing has 8 motors. It’s easy to imagine that it could have 8 independent batteries and could be able to perform a controlled descent if two fail. Similarly, it could have redundant sensors and use commodity “majority rules” logic circuits where all calculations are run three times. On top of that, they can already cut over to ground control if the computer can’t proceed.
Most of these are a total nightmare to implement. There is no such thing as a "controlled descent" into a dense urban center. Even keeping maneuverability with a single defective motor is a hard task, you need to develop and test for this.
>On top of that, they can already cut over to ground control if the computer can’t proceed.
So ground control can do what? The only reasonable way these could ever safely operate is with a high degree of sensor fusion, information about nearby crafts, their trajectories, obstacles, etc. combined together. You can not put a human into that system and have him make split-second decisions.
>With all that, it’s “just” a software problem.
Passenger plane automation is also "just" a software problem. It is just an enormously difficult software problem with exceedingly high stakes. The problem is not that it is "impossible", but that it is extremely difficult and not worth the risk getting it wrong.
Again, what is the acceptable failure rate for these things crashing into sky scrapers? If it is zero, we are very far away from them being able to take flight.
Anyway, detecting and avoiding obstacles should be in the menu. Maybe not as complex as at street level with people and cars doing unexpected things, but maybe with some added complexity that need to have into account like weather, inertia and things near landing sites.
I'm not sure you can avoid birds like a skateboad running onto the street.
If anything I expect dynamic avoid areas or timed "no fly" zones depending on bird mass. Maneuvering to avoid birds seems like a recipe for disaster.
If a drone flies at a speed that gives birds time to take notice and react/avoid, does it remove the danger ? I wonder if there is wide variation among species.
Self-driving car senses objects and must make difficult decisions: Is this radar reflection from stationary light pole or from a moving object? Is this a parked car or a moving car? Will this pedestrain/biker jump on the roadway? What if the car next to me going to turn left sharply?
No such thing exists in the air. The air is mostly empty, there are no stationary objects next to the route. If there is a building, aircraft can fly as far as 500 meters away from it - try finding a driveable road with no obstacles at that distance. There is no people or large animals. There is not even curb to hit.
Importantly, I bet there are no legacy vehicles in the approved routes - and I am sure they will have some sort of V2V tech to ensure that all objects in the air will transmit their position and intention.
Not every airplane does that. Commercial planes mostly do, but its not mandated under all circumstances as far as I know. (Speaking for my region only)
All commercial planes have it at this point. And most part 91 airplanes (General Aviation) have it at this point. Without ADSB you cannot fly in or near class A, B, or C airspace and must stay below 10,000 feet. Which largely makes having an airplane useless for most people.
There are still airplanes flying that have no electrical system but even then some of them have retrofitted a system.
I was answering with the purposes of flight-taxis in mind. I doubt that they will fly in high altitude and deem it quite likely that they might want to pass uncontrolled airspace at some point.
And if you're flying in uncontrolled airspace you just can't count on the presence of it. In fact, missing transponders are an occasional source of accidents in these spaces. There are lots of gliding enthusiasts that aren't equipped.
P.S.: Commercial planes (esp. big ones) should have hardly any contact points with flying taxis.
More degrees of freedom, easier collision avoidance. The mechanical part is a lot harder, of course, but the "self driving" part is a lot easier. We had autopilot and auto-land in planes well before we had anything of the sort in cars.
A self driving car needs to be able to reliably drive within ~1 foot of arbitrary obstacles, recognize junk in the road, recognize people in the road, obey vague hand signals from random cops, obey every weird traffic sign in the world, obey every traffic law, and disobey signs and laws when convention requires it.
Helicopters don’t have to deal with any of that. There won’t be any random obstacles. No human drivers to contend with. No conflict between legality and practice to navigate.
everything with a mass over 1kg+- in the air, is carrying a "gps" and something that is transmitting its location, and all air traffic world wide is(theoreticaly) assigned a flight path for each trip,witha specific flight level, ie: the planning is 3d, so the density per layer, never gets very high, and there are specific paths taken to change layers, there are never any dogs, or horses, or people trying to deek, there are no signs or maps, and in case of fully automated vtol
taxis, navigation will be to mm accuracy, with likely a routine to land fully autonomously on loosing the network, and a backup ballistic recovery parachute, if power is completly lost or another emergency situation occurs.
and also likely is that forward speeds will be low enough, that collisions with birds will be a low probability, and rotors will likely be schrouded, keeping the idiots heads attached when determinidly trying for that gota have selfie.
This is not correct. There is no hard requirement for GPS nor anything that requires “something that transmits its location”. There is also no requirement for a flight plan or communication with ATC.
I can take off from an uncontrolled airport with no GPS, no transponder, no plan, and no radio and as long as I’m in the correct airspace I’m completely legal and within my right to do so.
There are some differences with these rules through out the world.
This is just plain wrong. With glider planes, you can fly around wherever you like (in the non-restricted airspace) and aren’t required to have ADS-B or FLARM in Europe. Most people don’t do it but you could entirely rely on visual detection to prevent collisions.
I wouldn't trust anything that can't auto rotate or glide to be safe to fly. A drone style copter falls out of the sky if there is any failures at all at any point in time in flight making it an inherently unsafe method of travel.
> But the PRS-350 was released in 2010, so the fact that it still works at all at this point is exceptional.
The saddest tale about non-replaceable batteries. There is literally nothing inside that can break in the next century, except for one component, which is glued/welded/nailed in a way that destroy the whole device if you try to change it.
Okay, so being on this site, I imagine that you have seen a computer at least on TV - most likely you know the answer.
If you were Reddit, how would you try to recognize that an unwanted user is trying to register? Email address, IP address, browser fingerprint.
1. Use a brand new email address.
2. Use an IP to visit reddit that you haven't used yet. If you don't have such an IP, then it's time to go and get a new mobile subscription. Never use other IP to visit reddit from that account - it shouldn't be associated with any of the previously flagged addresses. You switch to wifi accidentally while being logged in, and that account is done. Also, the IP that you used to register is tainted too at the same time.
3. Use Firefox, just to be sure.
Of course there can other factors too, but especially the first two are the bare minimum.
I am sure you have not read more than the title ;-)
1. Done - so far - 12 new attempts, 12 new emails: 4 on gmail, 4 on protonmail, 2 on infomaniak and 2 on Tuta
2. Done - 12 diff exit points, in 12 diff countries, associated with each diff email (these also vary at times, by the intrinsic way the VPN picks dynamically servers in their exit points data centers). Used two diff VPN solutions: PIA and Mullvad (if that matters at all)
3. FF, Chrome, Brave, Safari (all on macos), Kiwi, Samsung Internet, Hermit and RIF (on Android) => changed the agent on a few, to match uniqueness of new email and new IP
Of course I read it. And unless you are posting pictures about yourself on reddit announcing that you are a ban evader, you are making mistakes that reveal your identity - you share one of your known IPs, or you share some of your registration details between the accounts.
Or alternatively reddit admins are psychics, in which case you should give up (or train to be a better psychic)
There are some machines that add some encrypted metadata to the image, which can be used as an attestation that the picture was taken with that machine, and it wasn't modified. E.g. search for Leica M11-P (Actually I have no idea if there are any others... if not, I guess there will be)
I agree that there is no magic bean, you have to learn stuff to make stuff.
But based on my experience, whenever someone says that something is "hard" or "complicated" or "difficult", it means 1 of 2 things:
Either the person doesn't understand it and assumes that no one else could understand it, or the person understands it, just wants to protect his/her job, and keeps everything a secret, forcing everyone else to reverse engineer his/her work. No exceptions. People in the first group also tend to solve these problems on the most complicated way, just to prove their point, until someone takes the time to actually look at the problem behind the solution...
>based on my experience, whenever someone says that something is "hard" or "complicated" or "difficult", it means 1 of 2 things: Either the person doesn't understand it and assumes that no one else could understand it, or the person understands it, just wants to protect his/her job, and keeps everything a secret, forcing everyone else to reverse engineer his/her work. No exceptions.
Not even remotely close to true. Some things are inherently hard, regardless of the motivation of somebody describing them as hard.
We have 80 years of experience building software, millions of bugs - and bugs costing millions, or billions, space explosions, medical equipment failures, national security compromises, and many other everyday issues, plus delayed projects with huge budgets, and going way past their budgets.
All this hard empirical evidence doesn't translate to "building software isn't hard, it's just because of ignorance/job security saying so".
Or, option 3, it is complicated and difficult. The fact that I understand something after 6 months of effort does not imply that it is simple or not hard.
The technically worst team I worked on was the one where no one could admit things are convoluted and difficult. Because team would rather reject that person as stupid then put an effort into simplifying messy complicated systems.
> People in the first group also tend to solve these problems on the most complicated way, just to prove their point, until someone takes the time to actually look at the problem behind the solution...
And that someone else will come in, say it is complicated and difficult ... and you wont believe them. You will thin they did not took time to understand it or they just want to protect their jobs.
reply