Google is Going to Eliminate Suicide Bombers — And That Is Not Necessarily a Good Thing
The Emergence of Autonomous Vehicles (Part 1)
Part 2 of this story can be found here.
HLSensory Overload: We’re Everywhere You’re Going To Be
If you ask the inventors of the suicide vest in Sri Lanka, or those who perfected it in Syria, they’ll tell you that it was because they didn’t have access to guided missiles. That is all about to change thanks to Google et al. Nearly all the technologies included in autonomous vehicles — from GPS to terrain recognition lidar were first used in cruise missiles and are now incorporated in high end military hunter-drones. Soon a driverless car can be packed with explosives, sent on its way to the NRA convention, and the terrorists can kick back, watch their Twitter feeds, and live on for the next attack. As you read this story, think of the possibilities…
Autonomous Vehicles: How This Stuff Works
We are already seeing the beginning steps of autonomy with driver assistance systems available today on new vehicles, like adaptable cruise control, lane keeping assist, braking assist, and self parking options. Within the next five to ten years, it is anticipated that autonomous vehicles will likely be available for purchase by the public. Nissan Corporations CEO Carlos Ghosn has established an aggressive goal of marketing the first semi-autonomous car in 2020. This technology is not limited to automobile manufactures either, as companies like Delphi and Google have been developing their own version of autonomous vehicles.
These vehicles have the potential to positively affect safety, decrease congestion, energy consumption, and offer viable alternatives for transportation to those with mobility challenges such as the physically disabled and senior citizens.
From business to pleasure to accessing health care and other societal goals, vehicles are a staple of daily life. In 2012, the highway system supported the generation of 15,685 billion dollars in gross domestic product. But all of these opportunities are offset by large costs imposed on users. In metropolitan areas time lost in traffic congestion results in lost productivity and wasted fuel consumption. For instance, a National Transportation statistic for 2011 indicated small urban areas with less than 500,000 populations resulted in 2.7 million gallons of wasted fuel.
The most telling statistic for review, however, is the deaths and injuries from crashes. In 2011, nationwide there were 5.6 million crashes resulting in 2.3 million injuries and 33,561 fatalities. Virginia’s statistics reveal 120,513 crashes resulting in 63,382 injuries and 764 fatalities for the identical period. The resultant losses in property, social costs, and civil litigation easily reach into the billions. It is believed that the use of autonomous technology can significantly reduce these alarming numbers
So How Does It Work?
The components necessary for an autonomous vehicle to function are varied depending on the desired level of automation required, but a general list would include the following:
Laser based sensing technology began in the 1970’s by the National Aeronautics and Space Administration (NASA) for space borne deployment. During the middle of the next decade, experimentation at Stuttgart University validated the use of lidar as a highly accurate method for topographic mapping. By the mid 1990’s laser based systems were being manufactured with the capacity for delivering sensors with the capacity to emit 25,000 pulses per second. The returns from the pulses were used to map features of topography with high accuracy. Today over 200 lidar systems are operational worldwide with the capacity to emit 250,000 pulses per second.
Lidar is able to map points in space using rotating laser beams that take more than a million measurements per second to form 3D models accurate to a centimeter. By preloading maps of traffic infrastructure, stationary items like signage, traffic lights and crosswalks are known. The lidar reveals moving objects like people and other moving traffic.
The development of radio detection and ranging has its origins in the United States Navy. Prior to radar deployment Navy ships could track other vessels by use of optics or sound ranging or primitive radio direction finding. In 1922, the U.S. Naval Research Laboratory made the first detection of a moving ship by use of radio waves. This radar principle was later refined during the 1930’s to include radio detection and ranging. The system proved invaluable during World War II and contributed to numerous naval victories.
Autonomous vehicles will rely on radar units mounted on the front and rear of the vehicle as well as the front and rear bumper. The units emit radio waves and measure the change in frequency of the return waves to provide range to objects in the vehicle’s environment. These units typically have a range of 60 to 200 meters (196 to 656 feet), and an associated beam width of 18 degrees to 56 degrees.
Global Positioning Systems — GPS
Most new vehicles come equipped with an on-board navigation system using GPS technology, and mobile units are available to consumers as add on equipment to older model vehicles. Portable electronic devices, such as cellular phones and tablets offer similar services.
GPS receivers usually track four to seven satellites at a time and couple the triangulated data with previously stored road map, and topographical data to display information in a user-friendly format. The GPS system provides free real time data in all weather conditions, and is available globally on a 24 hour a day basis. The receivers used to triangulate position are generally accurate to within 15 meters (49 feet) and newer models utilizing wide area augmentation system signals can narrow that margin of accuracy closer to 3 meters (9 feet).
While GPS systems provide accurate location information that autonomous vehicles will rely on, the system is not foolproof. The signals are transmitted via microwaves, which can be absorbed by water resulting in reception problems during periods of inclement weather. Additionally, line of sight to multiple satellites is required for functionality so heavy foliage and tall structures in an urban environment could hinder reception as well.
A variety of cameras may be mounted on the vehicle to validate and provide distance measurements to objects detected by the sensor network. The system may be equipped with still cameras or video cameras to capture images of the environment. These images would be compared with stored data in the on-board computer system to facilitate path planning and object detection. The range that these units will scan an area is typically between 100–200 meters (328 to 656 feet) in front of the vehicle with horizontal coverage between 30 to 60 degrees.
There are a variety of sensors currently housed within motor vehicles that provide indications of system performance and functionality. These internal sensors provide information on tire pressure, engine temperature, oil level and other system critical functions. Additional sensors are required for autonomous and connected vehicles to operate efficiently and will require real time updating of the environment external to the vehicle. This sensing of data will be compared to existing environmental data (stored mapping software) to allow for path planning and execution of vehicle maneuvers.
Sensing devices may be mounted at various levels around the vehicle to provide 360-degree horizontal and vertical coverage. These laser rangefinders are mounted at various heights to allow for horizontal estimations of object height as the vehicle approaches an obstacle. Vertical scanning provides details of the ground profile ahead of the vehicle. Ultrasonic rangefinders mounted on the sides of vehicles will provide side sensing and rear sensing capability.
The on-board computer system must be of sufficient computational strength to handle complex computing of all data received from the sensor network. Configurations are not standardized and will vary across the range of manufactured automobiles. Some systems may require initial input from the user through direct action or voice activated command to engage the system.
A fail-safe mechanism will be incorporated into the design to maximize safety in the event of system failure. Ensuring smooth transition between the operator and the vehicle is critical for safety and gaining user acceptance.
Should drivers misunderstand, misuse or otherwise become complacent as a result of overconfidence in the automated function, driving tasks could have dangerous consequences.