I got glimpses of a number of projects that I hope to share with you in the newsletter soon, but the one that really caught my attention is FogROS, which was just announced as part of the latest rollout of ROS (operating system of robot). Beyond a small name which is both a reference to the cloud element (fog/cloud – not to mention the fact that the new department has a breathtaking view of San Francisco and a frequent visitor, Karl) and a kitchen problematic French, there’s really compelling potential here.
I’ve thought a lot about the potential impact of cloud-based processing over the past few years, independent of my writing about robots. Specifically, a number of companies (Microsoft, Amazon, Google) have bet big on cloud gaming. What do you do when you’ve seemingly pushed a piece of hardware to its limit? If you have low enough latency, you can leverage remote servers to do the heavy lifting. It’s something that’s been tried for at least a decade, with varying effects.
Latency is, of course, a major factor in gaming, where a millisecond lag can dramatically impact the experience. I’m not entirely convinced the experience is where it should be, but it seems the technology has reached a point where off-board processing makes practical sense for robotics. You can currently play a console game on a smartphone with one of these services, so surely we can produce smaller, lighter and less expensive robots that rely on a remote server to perform resource-intensive tasks such as SLAM treatment.
The initial application will focus on AWS, with plans to reach additional services like Google Cloud and Microsoft Azure. Watch this place. There are plenty of reasons to be excited. Honestly, there’s a lot to be excited about robotics in general right now. It was one of the most fun weeks in recent memory.
Let’s start with the ExoHeal robotic rehabilitation gloves. The device, created by Saudi V Bionic, won this year’s Microsoft Imagine Cup. The Early Stage Team is part of a proud tradition of healthcare exoskeletons. In this case, it is an attempt to rehabilitate the hand following muscle and tendon injuries. Team leader Zain Samdani told TechCrunch:
The movement driven by the flexor linkage gives us the flexibility to individually actuate different parts of each finger (knuckles) while keeping the device portable. We are currently developing our production-ready prototype that uses a modular design to accommodate different patient hand sizes.
This is the third week in a row that Walmart has earned a mention here. First was funding from GreyOrange, which he partnered with in Canada. Last week, we noted a big expansion of the retail giant’s deal with warehouse automation company, Symbotic. Now it’s another big expansion of an existing deal – this time dealing with the company’s delivery ambitions.
Like Walmart’s work with robotics, the success of drone delivery has been… spotty, at best. Still, it’s apparently ready to put its money where its mouth is on this one, with a deal that brings DroneUp delivery to 34 locations in six US states. I quote myself here:
The retailer announced an investment in the 6-year startup late last year, following trial deliveries of COVID-19 test kits. The first trials were conducted in Bentonville, Arkansas. This year, Arizona, Florida, Texas and DroneUp’s native Virginia are added to the list. Once online, customers will be able to choose from tens of thousands of products, from Tylenol to hot dog buns, between 8 a.m. and 8 p.m.
There are even more question marks around this stuff than anything else, and I’ve long argued that drone delivery makes the most sense in remote and otherwise hard-to-reach areas. That’s why something like this Wingcopter deal is interesting. Over the next five years, the company plans to bring 12,000 of its fixed-wing drones to 49 countries in sub-Saharan Africa. It will cover places that have traditionally struggled with infrastructure issues that have made it difficult to deliver food and medical supplies through more traditional means.
“With the impending food crisis on the African continent triggered by the war in Ukraine, we see great potential and social impact that drone delivery networks can bring to people in all countries of sub-Saharan Africa by bringing food where it’s needed most,” CEO Tom Plümmer told TechCrunch. “Particularly in remote areas with weak infrastructure and areas that are additionally affected by droughts and other scourges, delivery drones of Wingcopter will build an airlift and deliver food from the sky on a hoist exactly where it’s needed.”
Legitimately exciting stuff, that.
In more cautiously optimistic news, Dyson released some interesting news this week, announcing that it has (and will continue) to pump big money into robotics research. Part of the deployment includes the redevelopment of an aircraft hangar at Hullavington Airfield, a former RAF station in Chippenham, Wiltshire, England, which the company purchased in 2016.
Some figures of the company:
Dyson is halfway through the biggest engineering recruitment campaign in its history. Two thousand people joined the tech company this year, 50% of whom are engineers, scientists and coders. Dyson is boosting its robotics ambitions by hiring 250 robotics engineers in disciplines including computer vision, machine learning, sensors and mechatronics, and plans to hire another 700 in robotics over the next few years. next five years. The master plan: to create the UK’s largest and most advanced robotics center at Hullavington Aerodrome and bring the technology into our homes by the end of the decade.
The main project highlighted is a robot arm with a number of attachments, including a vacuum cleaner and a human-like robot hand, which are designed to perform various household tasks. Dyson has some experience building robots, mostly through its vacuum cleaners, which rely on things like computer vision to navigate autonomously. Still, I say “cautiously optimistic” because I’ve seen many non-robotic companies pitch the tech as a vanity project. But I’m more than happy that Dyson changes his mind.
Hyundai, of course, has been pretty aggressive in its own robotics dreams, including its 2020 acquisition of Boston Dynamics. The automaker announced this week that part of its massive new $10 billion investment plans will include robotics, with the goal of bringing some of its most advanced concepts to market.
Another week, another big round for logistics/fulfillment robotics, as Polish company Nomagic raised $22 million to expand its offerings. The company’s main offering is a pick and place arm that can move and sort small goods. Khosla Ventures and Almaz Capital led the round, which also included the European Investment Bank, Hoxton Ventures, Capnamic Ventures, DN Capital and Manta Ray.
We finally reviewed Amazon’s limited-edition home robot, Astro, and Haje’s feelings were… mixed:
It was fun to have Astro roaming around my apartment for a few days, and most of the time I seemed to be using it as a roaming boom box that also has Alexa capabilities. It’s cute, and all, but $1,000 would buy Alexa devices for every conceivable surface in my bedroom and leave me with enough money to cover the house with cameras. I just keep wondering why Astro makes sense. But then, that’s true for any product that’s trying to carve out a whole new product category.
And finally, a little robot crab from Northwestern University. The little guy can be remotely controlled using lasers and is small enough to sit on the side of a penny. “Our technology allows for a variety of controlled movement modalities and can walk at an average speed of half its body length per second,” says lead researcher Yonggang Huang. “It’s very difficult to achieve on such a small scale for ground robots.”
Scuttle, don’t walk to subscribe to Actuator.