It sounds like the opening sequence of a sci-fi movie – the screen is black until suddenly it’s illuminated by a pair of bright eyes at the front of a self-driving car. While many of the movie portrayals of driverless cars can skew to the negative – from rogue pursuits to system meltdowns, the people creating the technologies in today’s world are striving to bring the positives imagined by science fiction into the future.
As we discussed in our recent post on the Verify blog and on the back of the inaugural Electronomous Conference (taking place in Ireland in 2018) looking at understanding of the emerging technologies in the areas of Advanced Driver Assistance Systems (ADAS), Connected and Electric Vehicles, autonomous automobiles are quickly becoming a reality.
In order for autonomous automobiles to deliver all the promised benefits safely, there is one crucial milestone to reach; they will all need to see the world around them instantaneously and accurately.
Vision systems for autonomous vehicles are proving to be one of the most challenging obstacles to overcome but as everything else aligns for autonomous vehicles, manufacturers are pushing for their development and looking to build solutions to overcome this last hurdle.
In addition to V2V communication and mapping technologies, cars will potentially be equipped with radar, cameras and LiDAR in order to enable them to see and react to the world. LiDAR; meaning Light Detection and Ranging, is the remote sensing method on which the majority of companies are hoping to accelerate the development of autonomous vehicles. LiDAR is fundamentally a distance technology – a remote sensing method that uses light in the form of a pulsed laser to measure ranges (variable distances) to the Earth.
The way that Lidar works is through light pulses that are sent out and reflected off nearby objects and are then detected by a sensor in the car. The relay time is used by a computer to map out the car’s surroundings in 3-D.
In addition to this, another piece of technology that is at play is the Inertial Measurement Movement which looks to account for the angle of the laser and the impact of the movement of the car. LiDAR sits between cameras and radar and can detect distance, objects and their shapes
Some notable companies like Tesla and AutoX are not using LiDAR in their current state due to possible flaws in the technology and the high costs associated, preferring cameras and radar instead. Last year however, Tesla’s stance on not using LiDAR in their autopilot mode of the Tesla Model S was questioned following a fatal collision made public in July of 2016. The autopilot mode of the Tesla Model S draws on a number of sensors including cameras, ultrasound and radar. But one piece of technology Tesla chose not to build into its cars was Lidar, which is in fact used by Google’s driverless cars.
Tesla CEO Elon Musk has, in the past, dismissed the need for Lidar, suggesting the technology “doesn’t make sense” in the context of a car.
“For full autonomy you’d really want to have a more comprehensive sensor suite and computer systems that are fail-proof. That said, I don’t think you need LiDAR. I think you can do this all with passive optical and then maybe one forward radar,” he said during a press conference last October.
However, despite Tesla’s stance on using LiDAR in their cars, work by companies like Velodyne, Quanergy, Oryx Vision and Luminar in developing a mass production cheap lidar sensor could be integral and push the development of autonomous cars to an even faster pace.
Most companies involved the the self-driving car space including Google-owned auto company Waymo are looking to build their vision systems around the technology.
It is not only small companies specifically focused on Lidar that are working in this space. Companies like Mobileye and Valeo are looking to build their own sensors having partnered with different car manufacturers like BMW. Additionally, companies like Waymo and Uber are working to develop self-driving LiDAR technology internally.
Waymo even filed a lawsuit against Uber, claiming the startup had misappropriated its trade secrets and patented information regarding LiDAR tech. Uber admitted in court filings in April however, that it still uses commercially available LiDAR systems in its self-driving vehicles because its in-house technology isn’t ready for the road. Acknowledging that its technology is not their own first-choice is a rare move, but as Uber faces a lawsuit from Waymo, claiming the ‘second best’ title might save its own autonomous automobile program.
Looking closer to home companies like Valeo Vision Systems, Movidius and ImageVision are competing in this space and working with leading brands while the Connaught Automotive Research Group are exploring all areas of assisted driving and autonomous capabilities and their implications.
From their initial start as Connaught Electronics, Valeo Vision Systems have been at the forefront of computer vision which led to the first ADAS project developed completely in Valeo Vision Systems to go to series production; Cross Traffic Alert with BMW in 2014. They are continuing to push the boundaries and solve the problems of autonomous driving.
What is clear is that the problem is yet to be solved but the companies are getting increasingly close. This technology could be the make or break for a lot of these companies as they increasingly share the other technologies necessary for fully autonomous vehicles. Cars may not be getting a pair of eyes attached to the front of them but they will certainly be able to see!
It’s safe to say that the LiDAR race is well and truly on!
Verify Recruitment is a specialist consultancy which recruits for technology organizations throughout Ireland, the UK and Europe. Founded in 2009, Verify Recruitment is the preferred consultancy of innovative technology clients, ranging from indigenous Irish start-ups to globally recognised brands.