In an abandoned car park in Pittsburgh, an unmanned car was driving slowly around the site. It had a name, called Terregator, who created it, 'self-identification and driving' Was used to describe the properties of the automatic driving car, so that this 'will roll the table' almost created history.
    The CMU R \u0026 D team named 'Terregator' car, there are six wheels, a few centimeters per second. A series of sensors, including sonar rings, cameras, and a single-line lidar rangefinder, are responsible for sensing obstructions Sensor and the environment. CMU on the birth of Terregator and become the birthplace of automatic driving technology, and this also opened the CMU on this basis for more than 30 years of technical research. Over the years, Terregator's creator, Whittaker, has been supporting the development of other automotive-related technologies.

   Now if there is an autopilot on the streets of the city, it is no longer so surprising, but the CMU and its Pittsburgh city, but because of the story of Terregator has become the oldest place to conceive unmanned technology. CMU's autopilot technology research and development began in the mid-1980s an opportunity. In that era, everything is from scratch. From the automatic driving car planning path, to the path tracking, to avoid obstacles, to find the right to achieve automation of the software system.
The challenge for autopilot is not just the obstacle itself, which means that the vehicle needs to perceive these obstacles and, more importantly, how the next step in the decision should be. There is too much work on the driving decisions Fuel Rail Pressure Sensor of unmanned vehicles. In other words, once the visual data is interpreted, then the vehicle must be on how to drive, how to turn to make judgments, in this part, is the place where innovative technology burst.

   Hebert recalled that when the CMU first started to develop autopilot technology, their car was only 'seen' for two days. At that time, the team used a camera-like sensor, with the image analysis technology, while the use of lidar direct access to 3D perception of information. These two methods of environmental perception are still being adopted so far, but have become more complicated than the 1980s. CMU's latest autopilot, a Cadillac SRX that debuted in 2013, was successfully held by Bill Schuster, a US Pennsylvania state legislator, at the airport. Technically, the vehicle is sensed by image analysis and measurement of road signs such as lane lines, while the vehicle has been able to achieve the average speed of the expressway. And really let it distinguish with the Terregator, in addition to speed and appearance, as well as fuel metering valve the dynamic things in the environment perception, whether people, cars, or bicycles.

   The next step in the study was to allow the vehicle to speculate on the movement of other vehicles, such as a car passing from it, as it allowed the autopilot to be sensed and traced from a simple surrounding environment to a dynamic response to the surrounding vehicle.

The results of these studies were demonstrated on the NavLab 5, which was subsequently developed by the CMU, a self-propelled vehicle based on the Pontiac movement. In 1995, the car was in the 'Cross-American Experiment' (NO Hands Across America, NOAA), from Pittsburgh, Pennsylvania to San Diego, California, traveled 4587 kilometers, of which 98.2% of the autonomous driving part.

   After that, in 2005, CMU took H1ghlander to participate in the DARPA Challenge that year, it was a modified general Hummer, it was on the road to a human driving vehicle overtaking. An autopilot drove over a human driving car, the 2007 DARPA City Challenge, and 11 of the more than 100 teams were out of the woods, including Boss - CMU based on Chevrolet Tai Hao's automatic driving modified car, and ultimately, it won the championship in the year, the average speed of 22.53km / h. At that time the track was 55 miles (89 km), the contestants were required to complete three driving tasks in the process and complete the race within 6 hours. CMU in the DARPA Challenge in the outstanding performance, directly to the United States General Motors Corporation donated 5 million US dollars to establish a second laboratory, automatic driving technology research and development laboratories, and by Rajkumar as co-director. Rajkumar also has an identity, founder of Ottomatika, an automated driving technology startup. In 2013, Ottomatika was founded and is committed to commercializing the results of CMU's autopilot society. Subsequently, the company worked with Delphi to carry out a number Temperature Sensor of cars in the city or highway driving active safety platform research and development. In 2015, Ottomatika was acquired by Delphi. Also in that year, Delphi with Ottomatika technology modified an Audi Q5 autopilot, from San Francisco to New York, self-driving time of 99%. CMU has already published 140 research results for autopilot technology, in which the Cadillac SRX has also become a brilliant addition, not only because of its groundbreaking completion of the depth of integration of many cutting-edge technology, but also because many sensors and kits have been The vehicle was embedded in the embedded.

   Boss's body carrying a bunch of equipment, roof laser radar, a lot of camera, the rear is also filled with electronic devices. There are two huge display in the car, the driver needs to constantly monitor the information, at the same time, even if there are three or four years old driver, must also know how to operate this car to drive the car, and to the 2011 Year, with the general company based on Cadillac modified, then the primary goal is to make the car look more normal.

In the SAE-defined Level 0-Level 5 autopilot level, Rajkumar said SPX reached Level 3.3. And for the real level 5, Rajkumar that at least 10 years, or even longer. Is the next generation of automatic driving to achieve the reserve, which is no longer just driving their own problems, but also involves many car gauge component development - in order Speed Sensor to truly achieve product, while taking into account the complex problems of various urban roads, Weather problems, but also the ultimate cost control.