Here is a magic road to the future – Sohu Technology

Here is a road to the future of Magic – Sohu technology on the road leading to the magic of unmanned, Intel forward. At the 1939 World’s fair, GM’s future world exhibition indicates that a new era of automotive control by embedded technology will eventually arrive. Early driverless car concept is more concerned about supporting the vehicle infrastructure, rather than the vehicle itself. At that time, it was hard to believe that driverless cars, but decades of technological progress and cross industry cooperation is making it a reality. According to the BI Intelligence report estimates that in 2020 there will be ten million cars on the road, HIS is expected to reach the driverless car sales in 2035 reached 21 million units. Let us return to this day. The sensor technology, computer processing power and machine learning (i.e. the ability of computer based on massive data accumulation, learning and creating mathematical algorithms) to create the Intel Jack Weast called "the perfect storm", namely unmanned. BMW, Intel and Mobileye announced the launch of BMW’s driverless car in the year of the shift to fully automated SAE international automotive automation is divided into five levels. Level 1 and level 2 still require the driver to hold the steering wheel ­ for example, adaptive cruise control system, responsive brake and parking assist system. The 3 level to achieve the car autopilot, but when necessary, the driver can recover control. On 4 grade driver needs less, passengers can even break. The 5 level is to achieve full automatic car, the car does not have a driver’s seat or control device. The realization of automatic or driverless cars need to deal with unimaginable massive data. Weast said that today’s mobile phone package contains several monthly GB data, including video and music streaming. In contrast, driverless cars will deal with TB data per hour. Too much data needs to be sent to the cloud, so the vehicle must have high performance computing capabilities to deal with massive data in real time. Driverless cars equipped with sensors, such as cameras, laser radar and radar, they can uniquely perceive the information surrounding the vehicle. For example, the camera may be able to see a person, but the radar can achieve depth perception, and found the difference between a real human and a human cardboard. The whole system must work together. "The car needs not only the high fidelity sensor, also the central brain intelligent high performance, it can deal with the virtual environment sensor data, create the body around the physical world, then make a steering and braking and other aspects of the very complex and rapid real time decision," said Weast. "There must be a closed loop in the car." Weast also explains that the autonomous part of the car comes from the ability to perceive situations: "in the real world, we need to recognize and understand many unique things, especially in complex urban environments. Cars need to be able to identify them accurately." With the capture of large data sets, data scientists can analyze all records相关的主题文章: