This is the fifteenth edition of AutonoMusings, which seems like a good milestone to collect some feedback from our readers.
If you have thirty seconds to spare, please fill out this very brief survey. We’d be grateful and you’ll be helping to steer AutonoMusings for its next 15 editions and beyond!
Microsoft has joined Eniac Ventures, and others, to fund Xwing’s pursuit of pilotless systems for helicopters and airplanes. Xwing is taking a “retrofit” approach to autonomous aviation as their software stacks are designed to also be installed on existing commercial aircraft.
There are two types of flyers in this world: Nervous Nellies, who irrationally believe turbulence will cause the jet to crash, and informed flyers, who know birds are the real threat.
An international collaboration that includes researchers from CalTech have created a drone that can herd an entire flock of birds away from the airspace near airports. The algorithm not only reduces the potential for collision, but can do so without disrupting the flock’s formation.
An Australia-based company, Resolute Mining, plans to rollout driverless trucks, loaders, and drills at its gold mine in Mali by the end of the year.
The company says local jobs will not disappear as it will focus on “up-skilling, training, and empowering the local workforce,” rather than seeking reduced labor costs. Whether or not that’s true will have an impact on how receptive other miners will be to the technology.
Kirstin Petersen, researcher at Cornell’s Collective Embodied Intelligence Lab, studied termite behavior hoping her observations might shape new algorithms for autonomous construction robots. She learned that some termites are natural-born leaders, others are hardworking go-getters, and some are woefully underperforming slackers.
It’s an insight that may have implications for swarm robotics, where the classic assumption has been that swarms consist of large numbers of identical units. Instead, a little individualism may go a long way.
Autonomy demands data. Lots and lots of data. For that data to be usable in computer models down the road (so to speak) a process called “labeling” is required. San Francisco-based Scale, which provides an API that accelerates the data-labeling process, just announced $18 million in funding led by Index Ventures, Accel, and Y Combinator.
Don Burnett, who left Uber’s self-driving truck program a few months before the axe fell on the project, teamed up with former venture capitalist Paz Eshel to form Kodiak Robotics in April this year. Now, the pair have just announced $40 million in Series A funding, which they will use to field test a fleet of autonomous big-rig trucks. Kodiak joins an increasingly crowded field in the AI-driven long-haul tractor-trailer space.
LIDARY is a deceptively difficult technology to get right. Companies often over-promise in a bid to gain investor and customer attention, and then end up under-delivering when they hit technical snags.
Solid-state LIDAR company Quanergy may have just pushed that game a little too far, drawing negative attention from news service Bloomberg. The article reports that both Quanergy’s investors and autonomous vehicle customers are growing frustrated with the company’s slow progress. Ouch.