Tuesday, August 20, 2013

Computer-Controlled Self Driven Cars

By Neeraj Mahajan

These are not exactly flying cars but the next innovation in vehicles will be even bigger surprise --- cars and buses –that drive on their own. You don’t even have to drive or keep your hands on the steering of these computer-controlled vehicles?

Since the 1950s the industry has been experimenting with various kinds of self-driving vehicles, some 40 years back General Motors came up with Firebird II designed to run on highway with an electrical wire embedded in the road.
Today’s vehicles have computers and sensors and are capable of driving safely while avoiding collision.  It is only a matter of time when these truly driverless vehicles using sensor-based technologies are able to communicate with other vehicles on the road to prevent accidents and have the ability to sense and respond to the surrounding infrastructure: stop signs, street lights and other traffic signals.
The future holds many interesting possibilities as far as buses are concerned. The urban transport planners are in particular excited about the prospects of regular buses being replaced by Self-Driving Vehicles or SDVs by 2022
The SDV or self-driving vehicles will replace every form of motorized transport - car, bus, taxi or light-rail. And by 2045, the last of the traffic signals will vanish because they will no longer be needed.
Researchers at the Institute of Electrical and Electronics Engineers (IEEE) predict that self-driven autonomous cars will make up 75 percent of vehicles on the road by 2040. All this will have huge ramifications on traffic infrastructure, personal mobility and attitudes.
Technology may not possibly change geometry but is certainly going to rewrite history!
Researchers at the University of California, Berkeley one of the preeminent universities in the world—which can boast of having produced 22 Nobel laureates and 4 Pulitzer Prize winners are field testing a bus that steers itself.
The self-steering bus developed by California Partners for Advanced Transit and Highways (PATH) follows magnetic strips embedded on the road. Although drivers may still handle acceleration and braking, the bus is capable of taking full control at any time. This technology could make life better for passengers by increasing efficiency.
The self-steering bus, which uses a camera mounted at the top of the windscreen to follow lines painted on the road, does not deviate more than a few millimetres from its course.
The magnetic guidance system also developed by UC Berkeley can both improve safety and provide a smoother ride for passengers. The system has the potential to follow bus rapid-transit bus-only lanes. 
Administered by the Institute of Transportation Studies PATH focuses on state-of-the-art traffic management, traveler information systems in addition to investigating new concepts and technologies for innovating, enhancing and improving transit solutions
The test used a 60-foot research bus traveling along a one-mile stretch of road near San Francisco. During the demo, the coach traveled in a perfectly straight line before pulling into a bus stop and stopping one centimeter from the curb.
Magnetic guidance technology also uses onboard sensors to track magnetic markers embedded every 1.2 meters (4 feet) on the center of the road. A computer aboard the bus reads the code generated by alternating polarity of magnets to determine the buses’ latitudinal and longitudinal position. A bus going at a speed of 60 kmph can process data from 88 feet of highway in less than one second.
PATH an acronym for cutting-edge transportation research stands for Partners for Advanced Transportation TecHnology, concerned with innovation through technology, rather than any specific mode of transportation.
PATH researchers have been experimenting with magnetic guidance systems for about 20 years and have used them to guide passenger cars but this is the first time they’ve applied the technology to buses on public roads. The need for all this is that with the rising cost of fuel more people need to get out of their cars, onto buses so as to reduce congestion.
Magnetic guidance has many passenger benefits. As the recent tests have shown, the buses need less time to load and unload passengers and thus can run with greater reliability and punctuality.  The technology’s also relatively cheap. As compared to a bus rapid transit system which costs about $273 million, magnetic guidance technology comes at a price of $5 million only.  Self-driving vehicles also dramatically reduces the number of parking spots. Here are some other likely effects on land use and commuting patterns:
Greater road density for traffic movement: Right now, a lot of road space is occupied by parking lots. This is a big problem in the metros where commuters have to park their vehicles at the nearest metro station and travel long distances to their place of work every day. Apart from this many high-density housing and business units have sprung up without proper parking space as a result much of the road space is occupied by parked cars. Self-driving cars can solve this problem. As this mode of travel picks up, even more land near apartments and offices will be available for driving.
Virtually no parking spaces: Self-driven cars will eliminate the need for any off-street parking. During peak hours virtually all cars will be driving people around. Late in the night cars could double- or triple-park in lane and still be instantly able to get out of the way if they’re blocking another car. So the only people who need off-street parking will be rich people who don’t mind spending extra to park their private car.
Potentially safer: Self-driving cars likely won't make human errors. Auto crashes typically claim around 38,000 lives per year. Over 80% of these are attributable to human error; negligence, distraction, incapacitation, malice or other uniquely human quality.
Fewer Accidents: Superior reaction time of self-driving cars, and their ability to warn each wirelessly about impending stops, means self-driving cars will be able to safely maintain smaller following distances than human drivers. This will increase the efficiency of each traffic lane, reduce congestion and make driving more attractive.
More nimble buses: Ask anyone they would say that the biggest problem with buses is the long wait waiting for them to arrive. The present-day buses are enormous, run infrequently and are   almost empty during off-peak hours. On the contrary Self Driving Buses will be cheaper to operate, so instead of a big bus stopping every 15 minutes, it’ll be feasible to have a smaller ones stop every 3 minutes. And because each trip will have fewer passengers, it will take lesser time to drop them to their destination.
Reduced labor costs: A self-driving car may not need many operators. What will be the need to employ thousands of people from conductors, ticket sellers etc when a computerized system can do everything and better?
The concept of a “bus” too will undergo drastic changes in the self-driving world. What many transport planner are trying to figure out is when so many low-cost options are available, why would anyone would want to ride a bus?
The answer to this question depends on which city you’re talking about. In smaller metro areas, falling cost of self-driving car/taxi service and reduction in congestion may causes many rail customers to switch to SDVs. As self-driving taxis become affordable even for poor commuters, smaller metro areas are likely to become even more car-focused and sprawling than they already are.
In larger metros, emergence of affordable taxi service may actually increase subway ridership. More suburban residents will take a taxi to the subway station and ride to work from there.  The greater efficiency of SDVs has the potential to increase the size and density of larger cities. In the process rail transit systems of large cities will seem more essential than ever.
If you define a "car" or as "a separate enclosed vehicle for every passenger", then  all cars, self-driving or not, miniaturized or not take vastly more space per passenger than effective public transit.  This may not be a problem in low-density suburbs, but in cities where there are relatively little space per person.  Self-driving cars may certainly improve the efficiency with which cars use the space but the bottom line will still be that if you want two crash-safe metal walls between every two strangers going down the same street, you will need a lot more space than if those two people can sit next to each other on public transport.
These days a driver who talks on the mobile phone may get rebuked for dangerous driving, but in the future multi-tasking while driving might be encouraged. By 2040 drivers of self-driving cars, would be reading, playing video games or surfing the web while the computer-controlled vehicles systems cruise around without any human input.
Google's experimental camera and sensor-packed autonomous car is legally roadworthy in the US. It has driven more than 48,000 kilo metres last year had only one minor accident that too when it was being driven by a human. A computer doesn't get distracted by passengers or the radio; it can simultaneously look at a map and drive; and there's little danger of it misjudging the speed or braking distance of either itself or other vehicles.
Already almost 1.24 million people are getting killed on the roads each year, and WHO estimates that car use will quadruple by 2050. Major carmakers, Nissan, Volvo, BMW and Audi, all see computer control as the next big step in improving safety and maximizing fuel-efficiency. Leading players like General Motors, Honeywell, Bosch, Daimler, Toyota, Siemens, Denso, Deere, Volkswagen and Google have unveiled autonomous car prototypes … the idea is to make cars that are more aware of their environment and giving them the on-board intelligence to help them make real-time decisions based on that awareness. It is possible in this manner to increase road safety, ease traffic congestion and minimize CO2 emissions.
Car and buses that 'think' will be really useful in big cities with traffic congestion, noise, pollution and high accident risk. At present there are 9,730 patent applications and claims pending from self-driving vehicles. The autonomous vehicles will be able to communicate online, with the central navigation systems in real-time to create the shortest journey route. The sensors on board the vehicles can download and upload traffic information to decide the best route to follow.
However it remains to be seen how much control the humans are ready to hand over. It may take some time but there is a scope for manual and autonomous cars to co-exist. Drivers who simply want to get from one destination to another in the shortest period of time will have the option of letting the car do all the work.
Already a number of cars have started to take over some of the driving tasks. Automatic braking when you get too close to the car in front of you is an example. Other examples include assisted parking, blind spot detection and cruise control. Automation for supporting the creation of better drivers, providing increasingly more capable cars that make driving safer - always keeping the drivers in control-- is the official line of Ford Motor Company.
The British perhaps have something different in mind. Oxford University's RobotCar UK project has seen a modified Nissan Leaf electric car fitted with a HK$1,150 navigation computer system that uses small, discreet cameras and lasers to create a 3-D map of its surroundings. Controlled by an iPad mounted on the dashboard, the robot can automatically drive a route it's already travelled, though it relinquishes all control the moment the human driver taps the brake pedal.
Smartphone apps that show vehicle speed, mileage, fluid levels and maintenance schedules seem well-suited to cars in which the "driver" suddenly has little need to look at the road, but a well-connected autonomous car could do much more, such as communicating with the owner's home. These could further be integrated with the electric devices like heating and cooking systems at home. They could be switched on… based on the proximity of the vehicle.
For instance air conditioning could be switched on or off triggered when the car is three blocks from home, likewise security lights can activated when the car drives away. Car parks, hotels and shops may also be able to reach out to nearby vehicles to advertise their proximity. All this is another reason why the driver of the vehicle does not have to actually drive.
The basic concept behind road-trains and platooning, is that a group of cars automatically move fast keeping a safe distance between them, allowing the rest of the road to traffic to flow freely. The essence of platooning is real-time wireless communication and processing.
The vehicles communicate with each other to maintain speeds and safe distances between two cars. There is real-time zero-latency connectivity between cars so that each vehicle is completely aware of what is going on with the other car. This helps them monitor and dynamically adjust speed limit.
The idea of the autonomous vehicle is not new. Autopilot technology has been used in aircrafts since 1912. The same is being used in cars with in-built computers. Google stands to make money from selling hardware and software for self-driving cars.
Another benefits of driverless cars, is that driverless cars could potentially cut down the traffic jams.
Self-driving cars are currently legal in Nevada, California, and Florida. Google is testing a "fleet" of driverless cars in these states.  
Self-driving cars are now legal on California’s chronically congested roads where car use radar and GPS tracking to drive in traffic flow to maintain a constant speed and reduce the gap between cars. But for now, as per Senate Bill 1298 legislation all self-driving cars in the state need to have a licensed driver in the front seat of the car. This is a precaution in case some unpredictable event that the creators of a car that drives itself were somehow unable to account for.
Self-driving cars present a thorny legal issue; who's to blame, the driver, automaker or developer of the technology in case a driverless car is involved in an accident? Clearly as self-driving cars become more common, many new laws will have to be written.
Metro Vancouver's transit authority is exploring the possibility of driverless buses guided by computerized maps and laser technology. TransLink is looking into driverless transit, which has already been implemented in the Netherlands.  
Driverless transportation has already hit the streets of the Netherlands and magnets in roads will soon guide buses in Shangahi.

Google is also testing 12 driverless cars in the United States that too accident-free.  Google’s self-driving cars have cameras on the top to look around them and computers to do the driving. Their safety record -- 300,000 miles ported that its cars without a single accident in really impressive.