Robotics

Last year was a quiet turning point in the use of flying robots in an unexpected space: Indoors. Increasingly sophisticated sensor and mapping technologies, machine learning, and engineering and materials science have combined to enable drones to operate inside — often overhead — in enterprise applications like inspection and inventory management.
The drones span sectors like oil and gas, which have long been adopters of drones for outdoor inspections of pipelines and infrastructure, but also retail and grocery. A company called Pensa, for example, makes drones to scan store shelves and track inventory. Verity Studios, founded by Kiva co-founder Raffaello D’Andrea, makes drones with failsafe algorithms to keep them stable in the event a propeller fails. That measure of safety has allowed Verity to put on extravagant indoor stage shows using drones as performers. Another company called Flyability makes drones for indoor inspection and exploration, and its Elios 2 drone, a confined space inspection UAV, has been a trailblazer in terms of case studies.
Bureau Veritas France, for example, used the Elios 2 for a ballast tank inspection on one of its cargo ship, resulting in a safer and faster inspection with high-quality HD images available post-mission. The inspection was performed with one drone pilot in three ten-minutes flights. Typically, the same inspection would require three to four inspectors utilizing ropes, oxygen monitoring devices, and extensive safety equipment. Flyability’s drone’s collision tolerance and onboard lighting allowed the pilot to maneuver the drone safely and efficiently without any external light source.
“[Crash tolerance] is the true enabler to gathering data in the intricate and hostile places where our customers are searching for insights,” says Patrick Thévoz, CEO of Flyability.
The company gets away with what is essentially a crash-resistant drone with a combination of onboard smarts and a physical cage. A previous model innovated a passive cage to protect the sensitive propellers from catastrophic collisions. Elios 2 adds to that passive mechanical protection with 7 vision stability sensors for GPS-free stabilization. By enabling the drone to function without GPS, the company removes one of the primary impediments to using most drone platforms, which rely on GPS for a number of navigation-related functions, indoors.
Safety is one of the big promises of confined space inspection using flying robots. Nuclear power plants, for example, can be dangerous worksites with highly radioactive areas that contain critical components. Traditionally, inspectors have very little time to inspect said areas because of the radiation dose, forcing them to run in and out to perform the inspection. For a thorough inspection, the plant needs to go into outage, which is extremely costly to the operator. 
Nuclear plants typically put reactors in outage in spring and in fall for maintenance, lower energy use periods due to generally mild weather. In spring 2020, over 50% of all US nuclear plants used ELIOS 2 for their inspection in high-dose areas, according to the company. In one particular instance, an operator was able to use the drone to find a steam leak and its cause without having to expose personnel to what would have been 3 Rem, which is 3x the amount of exposure needed in a year to develop cancer. In another instance, an operator was able to inspect a potentially critical flaw without having to go in outage, saving the site over $500K of downtime. In both instances, workers were able to perform the job without being exposed to radiation and without having to build and climb scaffolding.
Inspection has been a growth area for robots, with concepts ranging from rolling avatars to suction-cup drones designed to climb into tanks. DARPA’s recent SubT challenge accessed developers to push the limits of robotic controls and communications in confined spaces. The ability of drones to take the sky in tight spots lends increased flexibility, access, and speed, pushing various industries even further toward robotic adoption to complement human inspection regimens. More

As robots take on an increasingly diverse portfolio of roles, their charging regimen is going wireless. The same tech that enables charging pads to juice up your mobile device is also coming to robots, enabling them to return to base and get more juice without the complex docking procedures required of contact-based charging stations.
WiBotic, one of the companies leading the way in industrial wireless charging, which has a special focus on automation technologies like robots and drones, just got a big vote of confidence from the FCC, which granted equipment authorization for WiBotics’s high power transmitters and receivers, which provide up to 300 watts of wireless power. It’s the first time the FCC has granted approval of this sort of technology for use in mobile robots, drones, and other devices with larger batteries.
The coup for the company comes after a successful summer when WiBotic, based in Seattle, raised $5.7 million in Series A funding. The company’s wireless charging technology caters to the rapidly expanding automation sector, including aerial, mobile, and marine robots and drones. I wrote about the tech in 2018. Back then, Waypoint Robotics, which makes custom mobile robots for the supply chain industry, unveiled its EnZone Wireless Charging Dock, which was powered by WiBotic’s inductive charging technology.
“FCC approval is not only an accomplishment for our team but also for our customers and the industry,” says Ben Waters, WiBotic CEO. “Previously only low power cell phone and small electronics chargers or very high power electric vehicle chargers were approved for widespread use. WiBotic is now providing a solution that lets the entire automation industry take advantage of the wireless power revolution.”
Wireless charging is seen as a step toward a completely lights-out factory, one in which no human workers are required to intervene during standard operations. High power wireless charging systems that are strong enough to charge robot batteries at a distance of centimeters away from a transmitter enable charging with less human intervention and maintenance. 
“The FCC approval lets us meet customer demand by providing standard products to a rapidly growing industry,” said Waters. “As the industry continues to grow, robots and automation in general are facing more regulation and stricter safety and emissions requirements. We’re excited to help businesses solve some of these problems as they rapidly deploy larger autonomous fleets.”
WiBotic maintains that by optimizing power across a fleet, system efficiencies can be increased and maintenance costs can be reduced over time. More

Robots, drones, and now 3D printers. Automation is coming for the construction industry, and as I’ve written, the $15.5 trillion global industry is long overdue for an efficiency-minded technological intervention.
A startup called Mighty Buildings is taking the concept to the extreme plans to automate up to 80% of the building process in its factories using 3D printing. It claims it can produce structures with 95% fewer labor hours and twice as fast as conventional construction.
This is as much a story of materials science as it’s one about robots. The increased automation is thanks to a proprietary composite material used by Mighty Buildings’ 3D printers. The company can print more of the building structure than previously possible, including overhangs and roof structures. Robots help add interior and exterior finishes, while humans are left to do the detail work that’s still beyond the grasp of production-scale automation.
This sort of production line actually has a ready analog that’s been a success story for going on 70 years: It’s how cars are produced. The difference is that 3D printing reduces the need for standardization that burdens traditional automation, leaving room for design variation even within a production-minded operation.
This is hardly the first time 3D printing has been used for structures. In 2018, I wrote about researchers at the University of Nantes, which unveiled a prototype 3D-printed, robot-built home. The house, which took about 18 days to build, was made of a combination of polymer and concrete, which the Nantes researchers said would keep it insulated for decades.
But Mighty Buildings is novel in its extension of the concept to a commercial enterprise. By combining 3D printing and prefabrication, it’s created what seems to be a highly-automated, efficient, and design-friendly process.
“Because we’re building homes for people to live in, we’ve been very deliberate in carrying out our vision to make housing better. This isn’t software that can be debugged on the fly,” says Slava Solonitsyn, Mighty Buildings’ CEO and co-founder. “We’re now ready to scale our production with full confidence in our certifications and code compliance for both our material and technology.”
The innovative approach and efficiencies in terms of both time (a studio unit can be printed in under 24 hours) and money (the homes evidently cost up to 45% less than comparable stick-built houses) has attracted investment and accolades. The company graduated from Y Combinator in winter 2018 and has raised a total of $30 million from Khosla Ventures, Y Combinator, SV Angel, and CoreVC, among others. It’s also certified under California’s Factory Built Housing program and was the first company to achieve certification under the UL 3401 standard for evaluating 3D-printed building structures and assemblies. 
Now comes the hard part, which is bringing 3D printed homes to market successfully. So far, Mighty Buildings has installed its first two accessory dwelling units in San Ramon and San Diego, with additional units awaiting delivery. But the concept is promising, particularly as housing costs rise and large cities, including Los Angeles where I live, face a major housing crunch.
“With a strong foundation in robotics, manufacturing, and sustainability, the Mighty Buildings founding team knows the different facets of the issues that face modern housing,” says Eric Migicovsky, Partner at Y Combinator. “Accessory dwelling units are just the start in further building out their unique approach to building.” More

Through a pane of clear plastic, speaking through a mask, a checkout clerk at a grocery chain told my wife she was feeling sick yesterday. My wife asked (and I’m imagining her taking a big step back as she did) if she’d told her manager. The clerk replied she had but she was out of sick days and couldn’t afford to lose the pay.
That story is true, and it’s horrifying — both for the risk of outbreak it suggests and for the complicated labor realities it betrays. It’s also an anecdotal illustration of one more reason automation is coming to grocery stores, and fast.
Concerns of automation replacing jobs en masse are often overstated, but if you’re a grocery worker it really may be time to pay attention. The pandemic has only reinforced the drive toward automation in the brick-and-mortar grocery sector, which is increasingly adapting to an on-demand delivery paradigm while seeking ways to increase efficiencies and cut costs to compete with digitally native concepts. And, of course, increased risk of employee exposure and customer outbreaks at high-traffic stores have only spurred the automation conversation. 
Employees have quite naturally fought back against cost-cutting measures, including by unionizing (the store my wife visited is not unionized, which may contextualize the lack of good options for a sick employee), but at the same time chains large and small are increasingly deciding to pivot toward automation in what looks to be the first wave of a larger trend.
The latest example is Badger Technologies, a product division of Jabil, which has announced it is deploying multipurpose robots at Woodman’s Markets throughout Wisconsin and Illinois. The chain has 18 locations. 

Badger’s robots monitor inventory, part of a new generation of shelf-scanning robots. Though relatively new on the scene, shelf-scanning robots will become a common sight in physical retail locations within the next couple years. They roam stores and quickly scan merchandise on shelves, helping retailers keep track of inventory more efficiently than employees with scanning guns.
But their real benefit is in the data they collect, which cycles into purchasing and stocking decisions. By analyzing massive amounts of data, such as item popularity at a given time, retailers can gain something akin to e-commerce level insights into customer behavior and product trends. Badger’s robots will monitor product availability, verify prices, and deliver precise location data for more than 100,000 items at each location, part of Badger’s retail automation solution designed to elevate store execution, lower operational costs, and increase store profits.
“Most Woodman’s stores are over 240,000 square feet, nearly six times larger than the grocery industry average,” says Tim Rowland, CEO of Badger Technologies. “Not only can our robots perform shelf scans in hours instead of days, but they collect and connect critical data with the Woodman’s mobile shopping app to take customer experiences to the next level.”
The technology is impressive. According to the company, innovative imaging tools and neural networks detect out-of-stock items with more than 95 percent accuracy while identifying misplaced or mis-priced items, huge generators of waste and lost revenue, with over 90 percent accuracy.
To be sure, none of this is meant to directly cut down on human employees. Woodman’s in fact, is a 100-year-old employee-owned operation. But it’s easy to see how the drive toward efficiency paves the way for increasing penetration of automation down the road. According to an announcement from Badger, automating shelf scans will eliminate arduous manual tasks, especially given each store’s large floor space. 
As the pandemic rages on here in the U.S., automation has largely lost its stigma. The grocery sector will be a bellwether for larger changes coming across the service economy. More

Human-machine collaboration and accessibility have defined the first years of the robot revolution. If I’m placing bets, the hallmark of the next phase of robotic adoption will be dexterity, and a company that’s just exited stealth is my Exhibit A.
You probably know that robots are now prolific across a variety of industries that are relatively new to automation (materials handling, energy, logistics, biotech). But most robots are still designed to operate within carefully structured spaces where tasks are repeatable and variation limited. There are plenty of use cases that fit that bill, but for robots to become truly disruptive — to say nothing of a potential threat to skilled labor — they will need to be endowed with dexterity that allows for complex manipulation in unpredictable environments. 
We’re seeing lots of signs of development in that space. Just this month my colleague covered research out of the National University of Singapore directed at giving robots artificial skin to help them feel, a crucial step in enabling robots to manipulate objects they may never have encountered before. Soft robots have long been a leading area of study for just this reason, relying on compliant materials to emulate the soft touch a human has by nature.
Against this backdrop comes Dexterity Robots, an aptly named startup focusing on intelligent robots with human-like dexterity for logistics, warehousing, and supply chain — not coincidentally some of the fastest-growing consumers of automation technologies.
“While robots are the backbone of manufacturing, they have historically lacked the ability to adapt and operate in dynamic environments like warehouses,” explains Dexterity founder and CEO Samir Menon. “Dexterity’s intelligent robots constantly adapt to warehouse operations and do the tedious and strenuous tasks, which maximizes productivity by enabling humans to focus on meaningful work.”

They do that by focusing on human-like dexterity. Spinning off Menon’s Ph.D. thesis in Robotics from Stanford University, Dexterity was founded in 2017. Its core technology relies on Menon’s control theory framework describing how the human brain controls and coordinates the body, which, according to the company, can be translated to mathematical programs that control robots in a graceful human-like manner.
To achieve that level grace in motion, the company has a full-stack approach combining software and existing agile hardware. Its robotic stacks have powerful sensing suites that give them capabilities like touch perception, computer vision, force control, and contextual awareness, and the software and controls are tweaked to meet individual customers’ needs. Given that automation has grown of late largely by prioritizing task-agnostic hardware, this level of customization atop a modular, widely adaptable technology stack may start become more standard.
Dexterity is exiting stealth with deep pockets — the company has raised more than $56 million to date. It’s a sign of a progression beyond hardware for automation, which is now on the cusp of a new wave of adoption.
“Technically, Dexterity’s robotic solution can do what their predecessors could not. Their robot’s ability to learn as it picks, packs, and places novel objects is unsurpassed,” said Wen Hsieh, partner at Kleiner Perkins. “Dexterity also stands out because of their high-touch approach with customers, which includes gaining a deep understanding of customers’ needs, and then offering a Robots-as-a-Service offering. This unique pricing model allows Dexterity to deploy quickly and effectively, which results in an immediate performance and financial impact on customers’ warehouse operations.” More

The Marines are about to get their hands on an impressive bit of hardware: a wearable robotic exoskeleton that gives users super strength. The company delivering the unit, a defense-focused subsidiary of Sarcos Robotics developed the exoskeleton for industrial uses, including in energy and construction.
Still, in many ways this is a return to roots for Sarcos. In 2000, the company was part of a storied class of DARPA grant recipients working on powered exoskeletons for defense purposes. In many ways the XO, which conserves energy by remaining passive when not actuated, is the fulfillment of that research.
Another exoskeleton maker, Ekso Bionics, came out of the same DARPA grant.
According to Sarcos, the U.S. Marine Corps will test applications for its Guardian XO Alpha, which was first unveiled earlier this year at CES 2020, where it was named “Top Emerging Technology” by Digital Trends, “Best Robot” by PCMag.com, “The Best Ideas and Products of CES” by VentureBeat, and was recognized by WIRED Magazine as being one of the smartest technologies on the show floor. Although the suit may bring to mind nightmares of battlefield cyborgs, the more immediate applications will be in the realm of logistics, where heavy lifting is often necessary.
“The Sarcos Defense team is very pleased that the U.S. Marine Corps will be testing use cases for our Guardian XO Alpha version this year,” says Ben Wolff, CEO, Sarcos Defense. “Our military branches need to regularly address changing personnel issues and reduce the risk of injury from performing heavy-lifting tasks. We believe that our full-body, powered exoskeletons will be a huge benefit to the Marines as well as the U.S. Air Force, U.S. Navy and USSOCOM, who we are also working with on our exoskeleton technology.” 

Sarcos was founded as a spin-off of the University of Utah in 1983. For most of its existence, Sarcos has been grant-supported or has taken on unusual engineering challenges for anyone willing to fund applied science. Sarcos engineered the dancing fountain in front of the Bellagio in Las Vegas, as well as the animatronic dinosaurs for the Jurassic Park ride at Universal Studios.
More recently the company has moved hard into the commercializing its technologies, as I’ve written. The Marines will be testing use cases of the Guardian Alpha XO and will take delivery later this year. More

A leading online grocer is teaming up with a company pioneering the use of robots and small micro-fulfillment centers to bring ultra-fast grocery delivery to the nation’s capital. The partnership between Fresh Direct, which operates grocery deliveries in the northeast, and Fabric will rely on robotic micro-fulfillment to push the speed envelope in the red hot delivery race.
Early last year I wrote about the hazy state of the on-demand grocery delivery market, which has only recently emerged but by any measure will be enormous.

It’s hard to find a clear assessment of the impact online grocery delivery is having both on grocers and consumers. The online grocery paradigm is taking shape very quickly, but aside from hodgepodge delivery numbers and mind-boggling market reports predicting a market in excess of $100 billion, the state of the industry is a little uncertain.

Lockdowns and public-exposure concerns related to COVID-19 are only quickening the move to online grocery delivery as legacy players like Kroger team up with robotics firms and newer players like FreshDirect streamline their online-native operations. Headquartered in the Bronx, the grocer uses a hub-and-spoke operational structure and relies on data to move products from primary facilities to local facilities for end user delivery. 
Micro-fulfillment is an extension of the model. By placing automated fulfillment centers near end-consumers as opposed to in massive warehouses located outside cities, which how fulfillment currently works, Fabric believes it can deliver on a winning combination of speedy delivery, cost reduction, and scale of operations. 
“FreshDirect is the pioneer of grocery e-commerce in the United States,” said Elram Goren, Co-founder and CEO of Fabric. “Now with Fabric’s robotic micro-fulfillment technology, purpose-built for on-demand fulfillment, they are positioned to best serve their Washington D.C. customers profitably and at scale. We are thrilled to announce this strategic partnership as one of the first steps in Fabric’s continued expansion across the United States.”

In October 2018, Fabric launched the world’s smallest automated fulfillment center, and today the site processes up to 600 orders a day out of just 6,000 square feet, far smaller than most fulfillment centers. Fabric also recently announced the buildout of an automated grocery site with three temperature zones, the first of a 12-site deal in partnership with one of Israel’s largest grocery chains.
The company’s model of partnering with existing brands is now helping it gain a stronger foothold in the U.S., which is rapidly embracing grocery delivery. 
FreshDirect’s two-hour D.C. deliveries will go into effect later this year. More

The king of awe-inspiring viral videos, Boston Dynamics is continuing its push to commercialize its agile robot creations with a new pilot program at Ford. The car-maker will be leasing two robotic dogs, known as Spot, for its Van Dyke Transmission Plant.
The dogs will be on something of a short leash, guided by handler Paula Wiebelhaus, who uses a controller and will personally monitor the bots in operation. Painted bright yellow, the quadruped robots will each carry five cameras and two hours worth of battery power and will walk the floor capturing plant data and dimensions that will eventually be used to retool the line. In the future, the robots could be used to perform this task autonomously.
“We design and build the plant. After that, over the years, changes are made that rarely get documented,” says Mark Goderis, Ford’s digital engineering manager. “By having the robots scan our facility, we can see what it actually looks like now and build a new engineering model. That digital model is then used when we need to retool the plant for new products.”
Scanning physical spaces is a job well-suited to robots. Other robots — even other dog-shaped robots — are used for similar purposes. Aided by a partnership with Lenovo, UK-based React Robotics has a four-legged robotic helper called DogBot designed specifically for the construction sector, though with potential applications in other industries. Like spot, DogBot is a mobile sensor platform that can autonomously navigate spaces utilizing machine learning algorithms for locomotion, perception, and proprioception. 
According to Ford, the Spot deployments could save the company $300,000 per year, which is the cost of manually scanning the giant facility.

“We used to use a tripod, and we would walk around the facility stopping at different locations, each time standing around for five minutes waiting for the laser to scan,” Goderis recalls. “Scanning one plant could take two weeks. With Fluffy’s help, we are able to do it in half the time.” 
In line with larger sector trends, Boston Dynamics designed Spot to be application agnostic, outfitting it with baseline capabilities to perform multiple duties. As a result, the company is celebrating early deployments in a number of use cases. In one deployment, a construction firm in Canada used a Spot robot to automate the capture of thousands of images weekly on a 500,000 square foot building site, creating an ongoing record of progress and enabling the builders to identify growing problems and inefficiencies early.
Amusingly, Ford’s Spot will, at times, catch a lift atop another robot as it sits on its haunches like a robotic king. That small courier robot, known informally as Scouter, will ferry the dog robots up and down the aisles of the plant to conserve the dog’s batteries.  More