This has got to be the stupidest political ad I have ever seen
September 21st, 2012
|
The funny thing about this ad is that it makes me really want to support Jeff Brandes.
(of course, I’d have to look at where he stands on all major issues, but he got this one right)
First lets consider this concept we hear so much about called “creating jobs.” Based on this video you might be lead to believe that the way you “create jobs” is by ignoring all legislation that is not attached to a bill with the word “jobs” in it, or by doing the typical thing of pouring money into direct subsidies for businesses and going out to a lot of wind-bag photo ops on various white elephant projects.
In reality, jobs are created when an economy grows. Economic growth is spurred by innovation and improvements in business models and technology. Not only does this create jobs, but it’s what allows standards of living to improve and societies to grow in general. Every time someone invents something new or improves how a service or product is delivered, demand is created and economic growth occurs. Ultimately, it produces the need for upgrades, new business opportunities, demand for labor and returns on capital investments.
A modern economy cannot prosper by standing still. New ventures need to be created, products need to be improved and new technologies must be invented. That is what brings prosperity.
Therefore, anything that can be done legislatively to reduce the burdens that prevent innovation is a good thing and in the long run, will help create jobs and generally improve the economy.
How this relates to the driverless car:
The automated vehicle is a technology whose time has come. Existing technology is already very much capable of controlling a car’s speed, keeping it on the road and in the correct lane and avoiding hazards and obstacles. In fact, it can do this more reliably and safely than a human can, since computerized systems do not suffer fatigue or become distracted. Safe and reliable driverless car technology has been demonstrated for more than a decade. Most recently, Google’s experimental driverless cars have already logged 300,000 miles on public roads without a single accident.
The benefits to driverless cars are vast. Aside from increases safety, an automated driving system would liberate drivers from the mundane task of maintaining control of a vehicle, allowing them to read, surf the internet or take a nap. Because of the greater safety margins, it may be possible to allow driverless cars to travel at much greater speeds than manually controlled vehicles. They can also safely travel in groups, traveling at high speed with only inches between each vehicle. This can dramatically increase highway capacity and also reduce fuel consumption. Driverless cars do not slow down to gawk at accidents or other incidents, resulting in fewer ‘rubbernecking’ delays.
There are also numerous applications that driverless cars could be used for. Rental cars could be programed to drive themselves back to the rental agency, allowing customers to simply depart the car at an airport terminal and not have to bother returning it themselves. Individuals who cannot safely drive due to visual impairment or other disabilities, would no longer have to give up their autonomy and rely on others for rides. Long convoys of cargo-carrying trucks could be automated, providing savings in driver wages and fuel. Important deliveries could be made without the need to have a driver.
As is always the case with this kind of technology, there will be those who stand to lose their jobs due to changing needs for labor. It’s possible that vehicle automation could result in delivery and long distance truck drivers becoming less in demand or car services seeing a drop in business. Even so, it’s never valid to try to suppress a technology based solely on the fear that it will cause short-term job loses. New technologies may be disruptive of traditional labor trends, but they always end up creating more jobs by spurring more innovation and creating the need for upgrades. Such is the case with driverless vehicles, which have the potential to improve safety and efficiency and create entire new markets for things like in-car entertainment systems, mobile internet systems, automated traffic reporting and other products and services.
We will likely see driverless cars in our lifetimes, but it should be noted that the fully-functional driverless vehicle is not likely to come all at once. Already partially automated driving systems are available in production vehicles. Adaptive cruise control has become common on new cars. Modern vehicles now also feature lane-departure warning systems, some of which include the ability to guide a car back into a lane if the driver does not correct for drifting. A few cars can apply the breaks to avoid collisions. Even completely automated parking can now be found in production cars.
These features will only grow in availability and capabilities in the years go come. As such, vehicle automation will likely first arrive in the capacity of systems intended to assist drivers or actively control the vehicle to avoid collisions, but will not replace the driver. This may be followed by automated driving systems whose only function is to keep a car within a given lane on a highway. Eventually this will lead to cars that can drive themselves on highways and finally to general purpose driverless vehicles. It will happen, but it’s hard to say how long it will take to get there completely.
Why Jeff Brandes should be applauded for his stance on automated road vehicles:
The technology of driverless vehicles is very real and it will become commonplace. However, one of the biggest hurdles that must be addressed is not technical but rather regulatory. Road safety requires effective rules and regulations for how vehicles are operated and the laws that are currently on the books were written long before the possibility of driverless vehicles existed.
A few basic issues that need to be addressed:
- What are the design criteria for automated vehicle systems. What kind of redundancy and safety features are needed.
- When is it permissible to have full automation and when might it be not permissible.
- To what extent are competent licensed drivers necessary and what are their responsibilities when in the vehicle.
- If an accident does occur when a vehicle is operating under automation, how is fault assigned.
- What are the standards for signaling and communication between automated vehicles.
- To what extent are laws like speed limits and tailgating regulations applicable to automated vehicles.
- To what extent should the government tailor infrastructure to automated vehicles, such as creating special lanes or signaling for traffic lights.
These are important issues that need to be considered. It’s not unsafe to address these regulatory issues. In fact, it is a major enhancement to safety to assure that these kind of regulatory and safety considerations are properly addressed.
Driverless vehicles are not inherently unsafe. The level of technology may or may not permit them to operate on side streets and around pedestrians. It all depends on the level of sophistication and the circumstances. This is exactly what regulators should be considering.
I applaud Jeff Brandes for being a rare example of a legislator who is forward thinking and willing to address the changes technology demands of legislation. All too often, we find the government lags far behind developments in science and technology. That’s why we often find that old laws are shoe-horned into situations they were not intended to address. (For example, prosecuting internet crimes using ‘wire fraud’ laws that were intended to address transactions over teletype or using outdated intellectual property laws to govern things like software code and genetically modified organisms.)
We need more legislators with the kind of mentality of Mr. Brandes.
And no, it won’t mean killer robot cars will be mowing down old ladies!
This entry was posted on Friday, September 21st, 2012 at 3:33 pm and is filed under Bad Science, Culture, Just LAME, Politics, media. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.
View blog reactions
September 21st, 2012 at 6:59 pm
I enjoyed the article and believe that these innovations will be coming but it will be a slow process. Your comment on “killer robot cars” did remind me of a weekly TV show that I watched some 50 years ago (black and white) called Science Fiction Theatre. This one episode had a scientist who invented a self-driving car that accidently was released on the highway while they were still experimenting with it. The big concern was that although the car could see other cars, traffic lights and signs, it couldn’t see people (i.e. pedestrians). So the big race was to find the car and stop it before it reached a town or city. Fortunately it ran out of gas as it coasted up to the first stop light. Let’s hope that these newer designs have all of the bugs worked out of them before they go solo.
Quote Comment
September 21st, 2012 at 8:06 pm
I guess it was alleged that Google pushed the legislation, but that does not mean its bad, if it is otherwise valid law. And I agree that it’s nice to see a proactive rather than reactive government.
The view of a driver-less car on the side-street with no capability of driving in that condition is totally unrealistic and I can see how this would appeal to fear. Reminds me of the SNL commercial for robot insurance for old people. No surprise the narrator even sounds like an old lady. I think that is the demographic they were shooting for. People afraid of new things.
Quote Comment
September 22nd, 2012 at 4:35 am
Automated cars may also make things better for people who don’t use cars at all. Currently, city centers (which provide the greatest variety of shopping and other experiences, and are more likely to support independent businesses instead of just the big chains) are losing out to out-of-town strip malls and big box stores, because city center parking is too scarce and expensive. This makes life much more unpleasant for non-drivers, as out-of-town businesses are inherently harder to access by public transit.
This would no longer be a problem with automated cars, because they could drop people off in the city center and then drive by themselves to an out-of-town parking lot. It would also open up more space in the city center which was formerly used for car parking.
Quote Comment
September 22nd, 2012 at 6:34 am
Thank the Flying Spaghetti Monster for DVRs so a person can fast forward past all the inane political ads that come out around election season. When I do see an ad it usually has the opposite effect on me than what the opposing candidate or issue group intended, especially after doing my own fact checking.
The bill that Jeff Brandes and Joe Negron sponsored, and that was passed unanimously by the Florida legislature, was intended to allow automobile companies who are developing driverless vehicles to operate them on Florida streets and highways as a test bed. Manufacturers don’t see the technology being commmmercialized for at least 10 years but such business friendly bills are precisely what attract entrepreneurs and job creators.
Why granny is concerned about jobs anyway is a mystery. Her days on the assembly line peening rivets appear to be well behind her. An attack ad showing an unemployed taxi driver with a KITT car whizzing by would also be ignorant but would at least be less confusing.
Quote Comment
September 22nd, 2012 at 11:59 am
Blubba said:
In that case, it’s an even more potent example of a bill that can result in job creation. A company like Google is exactly the kind of company you’d want to be in your state as much as possible. Google has very very deep pockets, well paid employees and need for more locations.
If Google stops by your state to test their car, there will be a sizable group of test drivers, technicians, engineers and so on coming to do so. If they decide to set up shop there, it’s even better.
Remember, everyone who works for them needs to stay somewhere, eat and so on.
Hence, attracting Google is a perfect example of job creating policy. In this case, the best part is this is not one of those schemes to draw business to a state by throwing money at them and giving them subsidies. That is never effective in the long term and is very expensive. Here they are just accommodating them within the legal framework.
The idea that it is dangerous is absurd. No big company would expose themselves to the liability and bad press of putting a vehicle on the road without a driver and untested or unsafe hardware. Besides, Google doesn’t actually test their cars unoccupied. There is still someone there who could potentially take over.
Quote Comment
September 24th, 2012 at 12:20 pm
One issue that I’m sure is not insurmountable, but seems like a challenge nonetheless, is the transition point when a driver has to take control at the end of an automated route. Assuming that automated driving will be implemented first on freeways, long before they are implemented on surface streets, when the car drives itself to an off-ramp, the human driver will have to take over. What methods will ensure that the driver is prepared to take control and avoid an accident?
I can see a driver dozing off while his car navigates the freeway, with his newspaper draped across the steering wheel. The heavy sleeper snoozes right through the alarm that is meant to warn him that it will be his turn to take over. Does he suddenly wake up in a panic, unable to grip the steering wheel because of the newspaper?
A series of alarms should be able to warn the driver with plenty of time. A prescribed method of feedback from the human driver should be sufficient to tell the computer he is ready to take over. A lack of said feedback should probably result in the car coming to a stop before it enters an area it is unable to navigate. Like I said, it shouldn’t be insurmountable, but it does make me wonder.
Another issue that will arise when automation is implemented on freeways, but not on surface streets (and most likely even after) is that an increased throughput on the freeways will very quickly deliver more cars to the various destinations than those destinations have the capacity to handle, creating huge bottlenecks and massive congestion at the exits. It may very well be preferable to have what you might call such single-point congestion zones, versus distributed congestion all over the freeway system, but it will present real challenges.
Quote Comment
September 25th, 2012 at 8:56 pm
Shafe said:
One would assume that if the human driver did not respond at the end of the automated control period the car would be instructed to drive to a fail-safe stopping area. This would be relatively easy to implement.
Quote Comment
September 26th, 2012 at 8:22 am
DV82XL said:
Yeah, I kind of worked through that while writing my previous comment, and I think that should be simple enough if there are very few drivers who are not prepared to take control when appropriate. I spend a significant amount of time on Houston freeways, and that gives me ample opportunity to ponder how the human factor can fail. You have a whole spectrum including sleeping through the transition, remaining distracted by a smart phone even while indicating readiness to take control, being surprised by the sudden responsiveness of the car, and even taking control of the wheel but forgetting to operate the brakes. These may seem silly, but people are silly, dangerously so.
If the problem turns out to be very limited in scope and can be handled by providing shoulder space for a car or two at each exit, that’s fine. But if the human drivers prove inconsistent in their ability to take over, I would not be surprised if all transitions were forced to occur with the vehicle stopped. Obviously, that would create another set of traffic problems. But like many things, that may only hinge on one engineer’s elegant solution.
Quote Comment
September 26th, 2012 at 8:49 am
Shafe said:
This will be solved by the drive by wire system who will not let you perform dangerous operations but take over control, it would be an subset of the automatic driving system and might be the default setting on an car with it, think adaptive cruise control and automated lane guidance.
Guess that people getting used to the helper systems like this and get lazy will be an even larger problem.
Quote Comment
September 26th, 2012 at 12:04 pm
magne said:
It’s not just going to be laziness but the fact that too much automated driving will mean the loss of the skill set for driving in general. Unlike aircraft pilots that must continuously drill on simulators to stay sharp, automobile drives might go years between events where they might need to take control in an emergency, and are unlikely to stay in practice between times.
Quote Comment
September 26th, 2012 at 7:09 pm
I am none the less a little weary of automated cars on backroads. Highways are one thing and I’d be fine with the automated car there, but the challenges of areas that are not highways could present some unique problems.
For example, a deer by the side of the road. Those are well camouflaged anyway and the automated car’s machine vision needs to see it and realize that deer have a tendency to dart in front of cars. To make it even harder: it has to not be fooled by a holiday arrangement of Santa and reindeer.
A relatively simple task for humans but very complex for machine vision.
The other thing is other drivers. The car has to tell if a driver who does the wrong thing just made a single incident slipup or might be drunk and needs to be avoided.
Other unique and hard to predict hazards could come up. A tree that is hanging on power lines, but is not quite fallen, still not safe to drive under etc.
Like I said, highways are fine and possibly some major roads, especially if they are fitted with infrastructure sensors to warn and such. But there are a lot of strange and unexpected things on a twisting country road at night which could require some pretty high order cognition.
Quote Comment
September 27th, 2012 at 8:07 am
Traffic that is a mix of automated and human drivers would present more challenges. Drivers try to take advantage of what appear to be predictable events on the road. That includes trying to beat traffic lights, cutting off trucks that don’t tailgate, or darting around city buses. If human drivers begin to recognize regular driving behaviors of automated vehicles, you can be sure that they will try to predict them and take advantage to get ahead. I would imagine those drivers would be emboldened by the fact that the automated vehicle will not react emotionally or unsafely in an attempt to thwart the human driver. All that will create more difficult and unpredictable situations for the automatic vehicle.
Let it not be thought that I see these as reasons not to automate vehicles. But I think that until AI and machine vision have advanced sufficiently to deal with these challenges that they will remain limited to special purpose lanes and paired with humans.
Of course, fully automated driving could remain perfectly safe for tasks such as “valeting” themselves in fleet parking lots, as long as the general public is excluded from those areas.
Quote Comment
September 27th, 2012 at 3:26 pm
drbuzz0 said:
Detecting living animals (or people for that matter) is not that difficult if they were to a incorporate thermal camera(s). Using modern microbolometer cameras this shouldn’t be too expensive (especially if mass produced)
Overhead and side to side clearance is normally detected by using laser scanners. Not ideal (VERY processor intensive to process the data) but it works pretty well. (And again, not dependent on sunlight) Problem is, with all these types, driving towards a low sun you are practically blind. Even more blind than a human driver would be as cameras can’t squint.
Quote Comment
September 28th, 2012 at 4:02 am
This guy said:
Probably pretty easy to detect animals with thermal at lest in cold places. Main problem would be to differentiate different dangers, an moose at the side of the road is dangerous, somebody riding an horse is not. Even differentiating people and animals are hard, more so as people often walk in groups.
Road conditions is also an problem, speed bumps, potholes, snow are common.
Quote Comment
September 28th, 2012 at 11:50 am
This guy said:
Yeah, thermal cameras would help, although it would depend on the circumstances. They can be nearly useless for detecting living things if its hot enough or if there is enough clutter.
Some of it is common sense that just seems very hard to program to me. As mange says, a rider on a horse on the side of the road won’t dart in front of you, but a moose might. Similarly, a child running in their yard who was going toward the road might not stop and is worth slowing down a lot for. An adult who is briskly walking to check their mailbox is not really a concern. The body size is different, but might be hard to assess rapidly.
A person with their dog is less of a concern if the dog is on a leash.
Water is another hazard that can be difficult to judge. You don’t want to drive through a foot and a half of water, but a couple inches is ok. It can be hard to tell. It helps if you know the road and you can look around to see how high it comes up on the curb, if it is flowing pass the road and so on. Of course it helps if there are other vehicles going across it.
I don’t know that these are insurmountable, but I think they could be very difficult logical problems for a fully automated vehicle. They represent only very occasional events, but still are worth considering.
This is why I’m more comfortable with this kind of thing on divided highways, at least for now.
Quote Comment
September 30th, 2012 at 1:21 pm
One of the advantages of robocars is V2V communications. The first car to encounter an obstacle will relay the situation to other vehicles downstream. They can also automatically notify the central system about problems and be notified about problems. If there are animals and children in the neighborhood, then they can warn the passenger so they can pay attention and by ready to hit the emergency stop button.
Quote Comment