Not all that difficult. we're powering ours with human stupidity which is infinite, filtered thru a mesh of hashed bitcoin which are well known to be imaginary. The math -- which involves dividing stupidity by cellphone user intelligence (zero) shows that perpetual motion is not only possible, but inevitable. We'll be taking our product to market just as soon as we handle a couple of engineering glitches.
"I'm not convinced that Tesla will get to level 5 with their current hardware."
I think your analysis is perfectly valid. One thing though. Human drivers basically have two (at most) pretty good optical sensors with no useful range capability (human stereo vision only works out to about 6-7 meters). The sensors can scan right to left through about 180 degrees and up/down through 30-45 degrees. They are augmented by at most three very limited mirrors and maybe on modern cars by a flaky, small screen rear camera image when the vehicle is actually moving backwards
Yet on good days, some humans can do a very credible job of observing and interpreting their environment with that crummy sensor set.
Maybe automated driving really is mostly a software issue. Not that I believe for one second that Elon Musk can cause the necessary software to be written.
"Bitcoin and others are a drain on time, people, energy, and hardware resources. They provide little back."
Unlike a Gerald R. Ford class aircraft carrier (sticker price -- 8.5B USD) that gives one in return.... well... unh... wait a minute... I'll think of something.
Question: Is a nuclear aircraft carrier an asset? It costs money to maintain. No one you'd want to sell it to can afford to buy it. You can't use it to attack a serious enemy -- they'll sink it.
Haven't been to Detroit for close to 30 years, but three decades ago, no one in Detroit walked anywhere except from the house/office/store to/from their car. The chances of hitting a pedestrian are quite likely higher in the middle of the Gobi Desert than in Detroit. Especially during a snowstorm when everyone will be huddled around the TV seeking news of when the terrible calamity will come to an end.
"The skidding and traction problems have already been solved in the 90's."
I live in a place that gets 80 or 90 inches (a bit over 2 meters) of snow in an average Winter. Let me assure you that skidding and traction problems still abound although the switch from RWD to FWD vehicles helped a lot. 4WD/Traction control can help in getting going. The problem is stopping. ABS performs poorly on unpaved surfaces and even worse on snow and ice. Always has, and likely always will.
You're right about the sensors and figuring out where the road is. Figuring out where the road ends and the roadside ditch starts under a layer of snow can be non-trivial for a human familiar with the road. It's going to be a lot harder for the computers I think. GPS? -- with REALLY good maps... and differential correction... Maybe... But in worst case conditions in a canyon with cliffs blocking satellites and trees attenuating signal and multipath reflections. It's not easy for people and may well be even harder for hardware.
Boston can indeed get a lot of snow in a single coastal storm. And sometimes those storms queue up and after a while finding a place to put the snow becomes a real problem. OTOH, it rarely get bone chillingly cold there. Cleveland, Buffalo and Rochester get lake effect snows that can pile up pretty deeply. Montreal is, AFAICS the third coldest major city on the planet after Novosibirsk and Moscow. But many smaller cities -- notably Winnipeg -- are substantially colder.
Detroit (I've lived there) has nippy Winters and gets a bit of snow exacerbated by some of the least competent urban drivers on the planet. But it doesn't compare to any of the above for extreme Winter weather. Detroit is also flat as a pancake -- few curves, no hills -- which probably accounts for the incredibly awful handling characterstics of Detroit designed cars of the 1950s-1970s that drove people who lived in bumpier areas to buy cars made in Japan or Europe.
Anyway, it's a reasonable place to test cars, but one needs to remember that they will be used in far more demanding situations.
Complete, accurate, current maps would certainly help. But people have gotten by without them (mostly) for over a century. And I think autonomous cars are going to have to be able to do so as well. Otherwise every street line painting project, pothole repair, jack-knifed semi-trailer or new/altered traffic signal is going to throw them for a loop. One of the requirements for a serious autonomous vehicle would seem to be to be able to distinguish a road from a parking lot from a playground and handle each appropriately. They'll also need to be able to recognize and read street signs -- although that might require some sign standardization.
My bet is that open-street maps will be more than sufficient for mature autonomous vehicle technology.
There are BTW a few other, larger, problems that need to be solved e.g. non-standard traffic control devices (flashing red left turn arrows for example), "double intersections" where the traffic lights are easily confused with those for another intersection, etc, etc, etc.
"But I suppose that doesn't need an app. Not sexy enough."
OK then, how about we connect the lock to a IOT device that monitors all the traffic on your home net (and any unprotected wifi servers it can get to), packages up the data, and sends it back to our servers? It also unlocks the box if the delivery guy (or anyone else actually) waves a smart phone at it. You, however have to type a 17 character random one-time code into the device to get your stuff out of the box. Is that sexy enough to get funding?
Not really. I wouldn't be surprised that within a decade, you'll be able to snooze or answer email, or play video games on expressways with reasonable confidence that your car will not kill you or anybody else and will let you know if it encounters a situation where it needs help. That'd be useful for those of you that have and use expressways I think.
Cars that drive surface streets in LA at rush hour? Quite a bit longer. Boston? -- maybe never.
"If 'failure' means someone dies, then that's about two orders of magnitude worse than human drivers"
Worse than that I think. There are actually fewer than 100 traffic fatalities a day in the US. Which is still an awful lot.
I wonder if the advertising campaign isn't more to get their cars permitted to operate without a babysitter present than to persuade folks to actually get into them. On top of which, I think the initial applications will be either expressway driving (Good roads, many fewer special situations than surface streets) or people moving -- campus shuttles, airport shuttles. Travellng a preprogrammed route slowly without running into people or stuff.
"We SHOULD be aggressively pursuing practical energy storage solutions"
Of course we should. And we (humanity) actually are. Really. DARPA alone funded about 50 projects last year. Lots of researchers and engineers in universities and companies in the US, EU, Japan, China, etc are working on various "battery" projects. Lots of money to be made from better energy storage.
Trouble is that people are impatient and want decades worth of R&D done no later than the end of November at the very latest.
Then there are the folks who think no R&D is necessary because the problems are already solved. Trying to confront them with reality is like arguing with a Sportsfan, Jihadi or Libertarian.
Good question. Lot's of wave power projects announced in the past ten or fifteen years. Very few completed. Probably are a few in place and delivering power, but when I spent a few hours a couple of years ago searching for data on just one working wave power facility, I couldn't find one. Not one.
BTW, an Australian company called Oceanlinx did build a couple of fairly serious facilities a few years ago. Both of them sank. It's not clear they ever produced enough power to establish a track record, and the company went into receivership in 2014.
Yes. One thing though. Current laws and regulations assume a centralized power distribution system with power distributed through a "tree" to users at the terminal nodes. As we've found with attempts be Google to run fibre, getting access to any portion of the existing infrastructure can run into a lot of obstruction. It'll probably be greater with power which is physically dangerous and probably really does require some regulation to prevent electrocution of passersby, pets, kids, fire, police, etc, etc, etc.
A lot of good points there. My understanding (could be wrong) is that Lithium Ion packs of the type Musk is using require fairly elaborate computer management to isolate bad cells, prevent thermal runaway,and keep any cells from discharging completely -- bad for Lion I'm told. What are the odds that Tesla's hardware and software will work with generic or third party cells?
Could even be a legitimate issue -- especially if Tesla is modifying their cells to better suit the Powerwall application.
I think for 3.4 million people, you'd need several nuclear power plants. Remember that every few years, the plant goes down for many weeks for refueling and maintenance. We can't just ship everyone on the island to Manhattan until the plant comes back up. Also, the plants take a long time to build and recent projects in the US, Finland, France have had substantial problems with big time cost and schedule overruns.
We have underground utilities in the area I live in. They are reasonably reliable, but fixing them when they do fail is extraordinarily expensive and time consuming. They might be better in an urban setting if all utilities -- water, sewer, communications, power, whatever -- were run through tunnels big enough for humans or robots to work in.
Trouble is that, to paraphrase JM Keynes, the sun can not shine for longer than any practical, affordable, battery bank can hold out. For applications where you can't turn off the juice occasionally, you really need a nuclear/hydro/fossil fuel backup with wires connecting the generation to the users. And you need to pay to maintain that generation and distribution grid even if you don't use it all that often.
If you try to visualize a system using only wind/solar/waves and "batteries" -- as solar/wind advocates often do -- you'll end up with a system that doesn't always work. And by the time you've appended the stuff you need to make the lights come on reliably when folks throw their light switches, you're going to end up with very expensive electricity.
It's not that there aren't some applications, e.g. pumping water to "reservoirs", running air conditioning, where wind/solar can work fine today. It's that the high reliability and low costs US/Canadian electricity users are used to are VERY difficult to replicate with current "green" technology.
BTW - What's green about huge stacks of Lithium-ion batteries?
That's a bit of a problem. Not only was Puerto Rico broke and effectively in bankruptcy before Hurricane Maria, probably the brokest entity in the destitute island is the Puerto Rico power authority (prepa). It was $9B in debt BEFORE the power grid got ripped up.
Landlines per se weren't all that expensive. At least not in North America. Mostly about the same as today in current dollars. But non-local calls -- any call to places beyond walking distance -- were expensive. Long distance costs were outrageous. International calls were well beyond outrageous.
You can try downloading the hosts file from http://winhelp2002.mvps.org/ho... and installing it in the appropriate place -- e.g./etc/hosts on most(?) Unix distributions. It'll bust a lot of stuff of course, but if you can live with that, mostly ads will be a thing of the past. You may need to update the hosts file from time to time.
I'd suggest keeping a backup of the original hosts file (if any) and maybe of the last one you can live with as well.just in case the most recent one someday kills a site you can't live without.
Slashdot Mirror
"try making a perpetual motion machine"
Not all that difficult. we're powering ours with human stupidity which is infinite, filtered thru a mesh of hashed bitcoin which are well known to be imaginary. The math -- which involves dividing stupidity by cellphone user intelligence (zero) shows that perpetual motion is not only possible, but inevitable. We'll be taking our product to market just as soon as we handle a couple of engineering glitches.
"I'm not convinced that Tesla will get to level 5 with their current hardware."
I think your analysis is perfectly valid. One thing though. Human drivers basically have two (at most) pretty good optical sensors with no useful range capability (human stereo vision only works out to about 6-7 meters). The sensors can scan right to left through about 180 degrees and up/down through 30-45 degrees. They are augmented by at most three very limited mirrors and maybe on modern cars by a flaky, small screen rear camera image when the vehicle is actually moving backwards
Yet on good days, some humans can do a very credible job of observing and interpreting their environment with that crummy sensor set.
Maybe automated driving really is mostly a software issue. Not that I believe for one second that Elon Musk can cause the necessary software to be written.
"Bitcoin and others are a drain on time, people, energy, and hardware resources. They provide little back."
Unlike a Gerald R. Ford class aircraft carrier (sticker price -- 8.5B USD) that gives one in return .... well ... unh ... wait a minute ... I'll think of something.
Question: Is a nuclear aircraft carrier an asset? It costs money to maintain. No one you'd want to sell it to can afford to buy it. You can't use it to attack a serious enemy -- they'll sink it.
On top of wood being flammable, termites don't eat concrete ...just saying ...
Haven't been to Detroit for close to 30 years, but three decades ago, no one in Detroit walked anywhere except from the house/office/store to/from their car. The chances of hitting a pedestrian are quite likely higher in the middle of the Gobi Desert than in Detroit. Especially during a snowstorm when everyone will be huddled around the TV seeking news of when the terrible calamity will come to an end.
"The skidding and traction problems have already been solved in the 90's."
I live in a place that gets 80 or 90 inches (a bit over 2 meters) of snow in an average Winter. Let me assure you that skidding and traction problems still abound although the switch from RWD to FWD vehicles helped a lot. 4WD/Traction control can help in getting going. The problem is stopping. ABS performs poorly on unpaved surfaces and even worse on snow and ice. Always has, and likely always will.
You're right about the sensors and figuring out where the road is. Figuring out where the road ends and the roadside ditch starts under a layer of snow can be non-trivial for a human familiar with the road. It's going to be a lot harder for the computers I think. GPS? -- with REALLY good maps ... and differential correction ... Maybe ... But in worst case conditions in a canyon with cliffs blocking satellites and trees attenuating signal and multipath reflections. It's not easy for people and may well be even harder for hardware.
Of course we put beacons on the warheads. Do you have any idea how hard it is to hit a warhead that doesn't want to be hit?
Boston can indeed get a lot of snow in a single coastal storm. And sometimes those storms queue up and after a while finding a place to put the snow becomes a real problem. OTOH, it rarely get bone chillingly cold there. Cleveland, Buffalo and Rochester get lake effect snows that can pile up pretty deeply. Montreal is, AFAICS the third coldest major city on the planet after Novosibirsk and Moscow. But many smaller cities -- notably Winnipeg -- are substantially colder.
Detroit (I've lived there) has nippy Winters and gets a bit of snow exacerbated by some of the least competent urban drivers on the planet. But it doesn't compare to any of the above for extreme Winter weather. Detroit is also flat as a pancake -- few curves, no hills -- which probably accounts for the incredibly awful handling characterstics of Detroit designed cars of the 1950s-1970s that drove people who lived in bumpier areas to buy cars made in Japan or Europe.
Anyway, it's a reasonable place to test cars, but one needs to remember that they will be used in far more demanding situations.
Complete, accurate, current maps would certainly help. But people have gotten by without them (mostly) for over a century. And I think autonomous cars are going to have to be able to do so as well. Otherwise every street line painting project, pothole repair, jack-knifed semi-trailer or new/altered traffic signal is going to throw them for a loop. One of the requirements for a serious autonomous vehicle would seem to be to be able to distinguish a road from a parking lot from a playground and handle each appropriately. They'll also need to be able to recognize and read street signs -- although that might require some sign standardization.
My bet is that open-street maps will be more than sufficient for mature autonomous vehicle technology.
There are BTW a few other, larger, problems that need to be solved e.g. non-standard traffic control devices (flashing red left turn arrows for example), "double intersections" where the traffic lights are easily confused with those for another intersection, etc, etc, etc.
How the hell much profit is there for Google and Apple in that?
"But I suppose that doesn't need an app. Not sexy enough."
OK then, how about we connect the lock to a IOT device that monitors all the traffic on your home net (and any unprotected wifi servers it can get to), packages up the data, and sends it back to our servers? It also unlocks the box if the delivery guy (or anyone else actually) waves a smart phone at it. You, however have to type a 17 character random one-time code into the device to get your stuff out of the box. Is that sexy enough to get funding?
"Anything less is snakeoil."
Not really. I wouldn't be surprised that within a decade, you'll be able to snooze or answer email, or play video games on expressways with reasonable confidence that your car will not kill you or anybody else and will let you know if it encounters a situation where it needs help. That'd be useful for those of you that have and use expressways I think.
Cars that drive surface streets in LA at rush hour? Quite a bit longer. Boston? -- maybe never.
"If 'failure' means someone dies, then that's about two orders of magnitude worse than human drivers"
Worse than that I think. There are actually fewer than 100 traffic fatalities a day in the US. Which is still an awful lot.
I wonder if the advertising campaign isn't more to get their cars permitted to operate without a babysitter present than to persuade folks to actually get into them. On top of which, I think the initial applications will be either expressway driving (Good roads, many fewer special situations than surface streets) or people moving -- campus shuttles, airport shuttles. Travellng a preprogrammed route slowly without running into people or stuff.
"We SHOULD be aggressively pursuing practical energy storage solutions"
Of course we should. And we (humanity) actually are. Really. DARPA alone funded about 50 projects last year. Lots of researchers and engineers in universities and companies in the US, EU, Japan, China, etc are working on various "battery" projects. Lots of money to be made from better energy storage.
Trouble is that people are impatient and want decades worth of R&D done no later than the end of November at the very latest.
Then there are the folks who think no R&D is necessary because the problems are already solved. Trying to confront them with reality is like arguing with a Sportsfan, Jihadi or Libertarian.
Good question. Lot's of wave power projects announced in the past ten or fifteen years. Very few completed. Probably are a few in place and delivering power, but when I spent a few hours a couple of years ago searching for data on just one working wave power facility, I couldn't find one. Not one.
BTW, an Australian company called Oceanlinx did build a couple of fairly serious facilities a few years ago. Both of them sank. It's not clear they ever produced enough power to establish a track record, and the company went into receivership in 2014.
Yes. One thing though. Current laws and regulations assume a centralized power distribution system with power distributed through a "tree" to users at the terminal nodes. As we've found with attempts be Google to run fibre, getting access to any portion of the existing infrastructure can run into a lot of obstruction. It'll probably be greater with power which is physically dangerous and probably really does require some regulation to prevent electrocution of passersby, pets, kids, fire, police, etc, etc, etc.
A lot of good points there. My understanding (could be wrong) is that Lithium Ion packs of the type Musk is using require fairly elaborate computer management to isolate bad cells, prevent thermal runaway,and keep any cells from discharging completely -- bad for Lion I'm told. What are the odds that Tesla's hardware and software will work with generic or third party cells?
Could even be a legitimate issue -- especially if Tesla is modifying their cells to better suit the Powerwall application.
I think for 3.4 million people, you'd need several nuclear power plants. Remember that every few years, the plant goes down for many weeks for refueling and maintenance. We can't just ship everyone on the island to Manhattan until the plant comes back up. Also, the plants take a long time to build and recent projects in the US, Finland, France have had substantial problems with big time cost and schedule overruns.
We have underground utilities in the area I live in. They are reasonably reliable, but fixing them when they do fail is extraordinarily expensive and time consuming. They might be better in an urban setting if all utilities -- water, sewer, communications, power, whatever -- were run through tunnels big enough for humans or robots to work in.
Trouble is that, to paraphrase JM Keynes, the sun can not shine for longer than any practical, affordable, battery bank can hold out. For applications where you can't turn off the juice occasionally, you really need a nuclear/hydro/fossil fuel backup with wires connecting the generation to the users. And you need to pay to maintain that generation and distribution grid even if you don't use it all that often.
If you try to visualize a system using only wind/solar/waves and "batteries" -- as solar/wind advocates often do -- you'll end up with a system that doesn't always work. And by the time you've appended the stuff you need to make the lights come on reliably when folks throw their light switches, you're going to end up with very expensive electricity.
It's not that there aren't some applications, e.g. pumping water to "reservoirs", running air conditioning, where wind/solar can work fine today. It's that the high reliability and low costs US/Canadian electricity users are used to are VERY difficult to replicate with current "green" technology.
BTW - What's green about huge stacks of Lithium-ion batteries?
"... If enough money can be found"
That's a bit of a problem. Not only was Puerto Rico broke and effectively in bankruptcy before Hurricane Maria, probably the brokest entity in the destitute island is the Puerto Rico power authority (prepa). It was $9B in debt BEFORE the power grid got ripped up.
Landlines per se weren't all that expensive. At least not in North America. Mostly about the same as today in current dollars. But non-local calls -- any call to places beyond walking distance -- were expensive. Long distance costs were outrageous. International calls were well beyond outrageous.
You can try downloading the hosts file from http://winhelp2002.mvps.org/ho... and installing it in the appropriate place -- e.g. /etc/hosts on most(?) Unix distributions. It'll bust a lot of stuff of course, but if you can live with that, mostly ads will be a thing of the past. You may need to update the hosts file from time to time.
I'd suggest keeping a backup of the original hosts file (if any) and maybe of the last one you can live with as well.just in case the most recent one someday kills a site you can't live without.
You, sir, are no capitalist. Where would we be if everyone adopted your selfserving attitudes?
I think pilots (if they can find any) will discover that 10ft is LOTS of altitude if the rotors lock up for some reason.