Once the stuff of Hollywood fantasies, autonomous cars have finally become a concrete reality. Autonomous cars have started paving the way for a future where computing will completely take over the driving aspect of travelling by personal vehicles. Further, all such cars link to a central command system.
Light and Shade based “ShadowCam” from MIT
One of the most challenging grounds of research and development in autonomous driving technology is the preparedness or pre-calculation towards sudden or unexpected encounters with moving objects or people, straying into the path of self-driving vehicles. This creates ample doubts of the possibilities of collisions between such vehicles – in case of which the passengers of any of the involved vehicles, would be at the mercy of their driving software.
To tackle this possible safety hazard, many brains from numerous technological institutions and firms are burning candles at both ends to produce a sustainable solution, and the ones at Massachusetts Institute of technology seem to have come up with just the thing.
Called the Shadowcam, the MIT designed system functions to identify the smallest variations of the shadows surrounding the subject (autonomous vehicle) concerning their respective sources of light from all around to detect any possible approaching object – vehicle or pedestrian.
Using a sequence of video frame inputs from numerous cameras targeting specific areas such as floors or adjacent or opposite facing elements of the backdrop, Shadowcam detects changes in light intensity concerning time across every one of these frames to calculate if objects are moving towards or away from the vehicle. Generally, observations of such degree and detail are not a part of the average human being’s driving reflexes and training. By computing the received information and categorising each image according to the type of objects they capture – stationary or dynamic – the software programs a suitable action for the autonomous vehicle in the form of abrupt braking or gradual decrease in speed in case the object is approaching.
When compared with the existing LiDAR (Light detection and ranging) system, ShadowCam has proved to be 0.5 to 0.75 seconds faster than the former – a considerable window given the split-second nature of accidents.
Once perfected, Shadowcam technology can also be used in robots to help them avoid colliding with people and moving objects should either stray in their path. The research and development team at the institute aim to provide this technology with an X-ray like edge – to be able to detect speeding vehicles and fast-moving objects from behind obstacles and across blind spots. Currently, it has only been tested indoors and around parking lots, but research and development plan to push it through different settings for both indoor and outdoor lighting.
World’s first 5G driven autonomous vehicle testbed
Halfway across the globe from the birthplace of the Shadowcam, the world’s first autonomous car driving test ground is being set up in the South Korean capital of Seoul. The 5G based testbed will be the proving ground for future mobility technology and aims to determine whether the world is finally ready to move autonomously in everyday life.
Setup in the Sangam district of Western Seoul, the testbed is equipped with all the infrastructure vital to autonomous mobility including EV recharging stations, detailed three-dimensional maps of the roads and a control tower called the “future mobility centre”. The testbed is aimed at putting on trial, the Cooperative Intelligent Transport System services which will be designed to communicate via the new 5th generation wireless communication network. Superfast connection speed, multiple device connectivity and low latency are a few of the networks advantages and when coupled with the detailed imagery of streets and travel grids will possibly make autonomous driving a practical reality much sooner than its timeline.
South Korea has come to be a pioneer in the use of 5th Gen services and aims to connect more than eighty neighbouring cities to this network by the end of this year. Four buses and three cars full of passengers have been set up to ferry people through traffic and recognise signals while doing so to acknowledge and move around obstacles on the road.
The country also has a fully operational testbed in Hwaseong district, which has not been opened to the public. Located approximately 60 kilometres away from the city of Seoul, this testbed is limited to being used for vehicle testing and certification examinations.
The world is fast-changing, and though such advancements are still a few years away for the Indian mobility system, the excitement that its finally happening somewhere in the world is hard to contain.