The secret to self-driving in the snow could be hiding underground

28 Feb 2020

Image: © magdal3na/Stock.adobe.com

This week in future tech, researchers are tackling the challenges of all-weather self-driving cars, touchy robots and better energy sources for a low-carbon future.

One of the challenges facing the development of self-driving cars is accounting for all kinds of weather. In the snow, for example, cameras can no longer recognise lane markings and traffic signs. And even in heavy rain, the lasers of LIDAR sensors can malfunction.

But a team from MIT’s Computer Science and Artificial Intelligence Lab (CSAIL) thinks localising ground-penetrating radar (LGPR), developed at the MIT Lincoln Laboratory, could be the answer to this problem.

This technology sends electromagnetic pulses underground to measure an area’s specific combination of soil, rocks and roots. It creates a unique mapping fingerprint of sorts that a self-driving car can later use to localise itself when it returns to that particular plot of land – no cameras or lasers required. What’s more, these subsurface maps will be relevant for longer as underground features change less frequently than overground ones.

Image: MIT CSAIL

Testing in snowy conditions, the LGPR system’s average margin of error was about an inch compared to clear weather. However, the system was off by about 5.5 inches in rainy conditions. It’s believed this is because water soaking into the ground creates a larger disparity between the original mapped LGPR reading and the current condition of the soil.

Further work is needed before LGPR is fully road ready, not least of all because the current hardware is six feet wide, so not yet small and compact enough to be integrated into commercial self-driving vehicles.

A new paper about the project will be published in IEEE Robotics and Automation Letters and will also be presented in May at the International Conference on Robotics and Automation in Paris.

Micro-hydro power

Belgian scale-up Turbulent has developed micro-hydro vortex turbine units that can generate electricity from rivers and waterfalls with a height difference of just 1.5 to five metres. A unit installed for the Green School in the Bali jungle provides as much renewable energy as four tennis courts of solar panels, servicing the hundreds of students and staff at the school as well as its local neighbours.

“It is also possible to connect a whole series of turbines together, in order to service larger regions,” said Turbulent co-founder and CEO Jasper Verreydt. “What used to require a dam, for instance, can now be achieved with a series of hydro turbines. The resulting energy output is the same but they are much faster and cheaper to install and have far less impact on the environment.”

A water vortex in Bali, Indonesia.

A water vortex in Bali. Image: Turbulent

There are also farming communities in Chile using Turbulent’s hydropower units to extract energy out of irrigation canals. New installations in Taiwan, the Philippines, Suriname, France and Estonia are in the pipeline, and Verreydt said there are millions of small rivers and waterfalls suited to micro-hydro power plants. In Europe, the company has identified more 350,000 suitable sites.

Robotic fingers

In a paper published to IEEE/ASME Transactions on Mechatronics, researchers at Columbia Engineering introduced a new type of robotic finger with a highly precise sense of touch. They achieved this with a novel approach using light to sense touch.

Under the ‘skin’ of the robotic finger is a layer of transparent silicone, into which scientists shined light from more than 30 LEDs. The finger also has more than 30 photodiodes that measure how the light bounces around. So, whenever the finger touches something, its skin deforms and the light shifts around.

These changes can be measured to extract information about the contact made and, since light can bounce around in a curved space, force detection over complex 3D surfaces is possible.

Though it collects almost 1,000 signals, this robotic finger only needs a 14-wire cable to connect it to a hand, and no complex off-board electronics. For the next stage of testing, the researchers will install their fingers on two robotic hands capable of grasping and object manipulation, to see if they can try and demonstrate dexterous manipulation abilities based on tactile and proprioceptive data.

Ireland’s hydrogen opportunity

Hydrogen fuel generated using renewable energy sources could be a game changer for a low-carbon economy, and a new study published in Nature Energy identifies processes that could help this country hit its climate targets in transport, heating, energy and storage.

Hydrogen can be produced from water through electrolysis, but current state-of-the-art electrolysis requires support-associated processes such as desalination of water sources and further purification and transportation.

This study examines how we could design electrolysers (devices that split water into hydrogen and oxygen using electrical energy) suitable for producing hydrogen from low-grade and saline water, with common impurities such as metal ions, chloride and bio-organisms. Not only is this a more abundant resource, it also means that potable, clean water – a precious resource – wouldn’t have to be used in hydrogen production.

NUI Galway’s Dr Pau Farràs, the lead author of the study, is one of Ireland’s experts in synthetic chemistry and catalysis. He said: “With the right vision and drive, Ireland can be a world leader in developing new clean technologies as it doesn’t have other major energy industries to displace. Hydrogen can deliver to all sectors of society, creating a zero-emission economy.”

Want stories like this and more direct to your inbox? Sign up for Tech Trends, Silicon Republic’s weekly digest of need-to-know tech news.

Elaine Burke is the host of For Tech’s Sake, a co-production from Silicon Republic and The HeadStuff Podcast Network. She was previously the editor of Silicon Republic.

editorial@siliconrepublic.com