Robotics

Badger Technologies
Shelf-scanning robots, a technology that debuted in major grocery chains, are now venturing out of the dairy aisle and into home improvement. Automation developer Badger Technologies recently announced it’s bringing its scanning robots to Busy Beaver Building Centers, popular in the midwest, to monitor on-shelf product availability and verify prices for more than 30,000 SKUs.”Badger Technologies’ retail automation solutions close important data gaps that prevent retailers from getting an accurate picture of shelf conditions, buying trends and customer preferences,” says William (BJ) Santiago, CEO of Badger Technologies. “We complete storewide scans in hours, not days, and offer instant access to intelligent data to elevate shopping experiences and store profitability.”This is an important milestone for a technology that’s primarily been marketed for grocery retailers but that developers hope will spread to brick-and-mortar retailers of various kinds. The technology class received a bit of a PR backslide when Walmart killed a large contract with another player in the space, Bossa Nova.The Walmart contract notwithstanding, shelf-scanning technology has emerged as a crucial way for brick-and-mortar to adopt some of the efficiencies of ecommerce, reducing waste and extracting valuable insights about product performance. A number of developers in addition to Badger Technologies have commercialized shelf-scanning robots, which are typically automated ground robots that navigate retail spaces autonomously (although one company has created a shelf-scanning UAV).The data that these robots collect while roving aisles at retail locations, combined with powerful analysis, is meant to increase efficiency by solving for the $1.75 trillion “ghost economy,” defined by out of stocks, inaccurate price execution, and lack of product location optimization industry-wide. With up-to-date inventory information, managers using robots like Simbe’s flagship Tally robot can enact faster operational decisions at the store level and more nimble inventory management. The problem of poor stock management is so pervasive that inventory mishaps account for more lost revenue than theft. Brick and mortar retailers and especially large chains are doing all they can to shortcut the traditional inefficiencies of operating a storefront. Granular data models of how inventory moves through a store are considered a holy grail in that quest, bringing e-commerce analytics to traditional retail. Retailers, including smaller regional chains, are responding.”Technology is a key imperative in our mission to be the favorite neighborhood home improvement center in the communities we serve,” says Adam Gunnett, director of IT for Busy Beaver Building Centers. “We relentlessly look for innovative ways to increase efficiencies and empower our associates to provide legendary customer service, which is why the pilot with Badger Technologies is so exciting. We expect to demonstrate how Busy Beaver can keep our shelves fully stocked with the right mix of products—at the right prices—without overburdening our dedicated team members.”

Like many stories about the intersection of technology and retail, this one has a pandemic bent. Home improvement and DIY have boomed as consumers spend more time at home. According to the Joint Center for Housing Studies of Harvard University, spending on home improvements grew by more than 3% last year to reach nearly $420 billion, despite the fact that the U.S. economy shrank 3.5% during that same period. Busy Beaver is smart to use the good times to invest in technology that could save it money in the long run. More

Skyline
Window cleaners are a common sight in New York City, where they work high above the sidewalk dexterously cleaning skyscraper windows. It’s a transfixing sight, and it’s also an incredibly dangerous job. It’s also one that may soon fall to robots. That’s thanks to a new agreement between the developers of a window washing robot named Ozmo and Platinum, a building maintenance service provider in New York.  Human window washers are transfixing to watch in action, but the work is incredibly dangerous. During one 15 year period, OSHA tracked 88 window washing accidents, a full 62 of which resulted in fatalities. That grim statistic highlights the thin margin of error when working sometimes hundreds of feet above the ground.Automation can address the safety concerns and lead to greater efficiency in a task that hasn’t had a substantial technological update in decades. The use case makes a lot of sense: managers of every commerce building in major U.S. cities need some way to clean their building’s exterior glass. In fact, window cleaning is $40B global market. The surfaces involved are generally uniform and the path predictable. It’s a perfect recipe for successful robotics development with a proven customer base.

[embedded content]

“Facade maintenance is integral to the health and spirit of a building,” says Michael Brown, CEO & chairman, Skyline Robotics, maker of Ozmo. “Automation is poised to play a key role in the future of façade access work as it will both increase efficiency and reduce risk, and this is just the beginning.”The system combines artificial intelligence, machine learning, and computer vision with a robotic arm designed by KUKA, a leader in industrial robots. Ozmo uses a force sensor and knows how fragile glass is, and AI helps the system remain stable, even in gusty conditions. The system utilizes lidar to scan a building’s facade, memorizing surfaces and planning a cleaning path, which it continually updates.”The application puts several of the latest advancements in robotics to work in a new and exciting business sector and brings efficiency and safety to building maintenance,” says John Bubnikovich, North America chief regional officer for KUKA Robotics. “Advancements that made such a daunting task possible include on-the-fly control of the KUKA robots in terms of pressure applied during the cleaning process as well as the ability to compensate for a moving scaffolding due to weather and changing architectural building features.”

Platinum will add Ozmo operations to its existing window cleaning division, Palladium Window Solutions. Skyline Robotics, which recently secured $6M in funding, will train the Platinum staff and certify them as Ozmo operators. The Platinum staff will then run the operations of the Ozmo system.”Platinum’s commitment to driving innovation and adopting new technology is paramount to our market-leading success,” said James Halpin, CEO, Platinum. “Thus Skyline, whose DNA is based on innovation, is an ideal partner as its cutting-edge technology will help us further increase our market share and remain one step ahead of the competition.” More

A key part of realizing the future of commercial drones will be drone traffic management: An integrated way to manage airspace for UAV. That’s the goal of a recent trial in Japan led by NEDO (National Institute of New Energy and Industrial Technology Development Organization) to develop a drone traffic management system for multiple drone operators to fly in the same airspace safely.The trial, closely watched in the industry, brings together several prominent companies and consortiums, including ANRA Technologies, BIRD INITIATIVE, NEC Corporation, All Nippon Airways (ANA), and other partners. It will take place above Wakkanai City in Japan using ANRA’s airspace and delivery management software platforms. The testbed is part of an ongoing R&D effort led by NEDO with the aim of integrating drone traffic management and creating a blueprint for a nationwide traffic management system. Future use cases include drone-based logistics, disaster response, and inspection.In the U.S., non-military drones are not integrated into a nationwide air traffic management system except under voluntary programs. Currently, the technology used to keep track of commercial and military aircraft does not account for drone traffic. However, as in Japan, efforts to evolve the national airspace management ecosystem are underway. For example, a company called AirMap has been aggressively lobbying for a comprehensive drone aerial management service. Its technology ID’s individual drones, enabling tracking and paving the way for the kind of nationwide net used to monitor commercial and military aircraft. Raytheon, which makes the technology that powers the existing traffic management system for crewed aircraft in the U.S., and AirMap are hoping to develop solutions to provide a complete, real-time view of manned and unmanned flights in U.S. airspace, helping allay rising fears that increasing drone traffic is putting the national air transportation network at grave risk. There have already been a handful of verified aircraft collisions with drones, and several more near-misses, raising awareness of the growing problem.In Japan, ANRA’s decentralized traffic management platform will help to coordinate and negotiate the airspace used between operators to avoid collisions between drones. The project utilizes automatic negotiation AI technology, which is being researched and developed with RIKEN and Industrial Technology Research Institute, and digital twin technology, which is being researched with the National Institute of Informatics.One of the test cases will be pharmaceutical delivery via drone. The results of the test will be shared over the next few months. More

University of Zurich
Expert human drone pilots have proven incredibly adept at piloting UAV through complex courses at high speeds which are still unmatched by autonomous systems. But researchers at the University of Zurich and Intel Labs are collaborating to change that, and their work, recently presented in the journal Science Robotics, could have far-reaching implications for the future of commercial drones.”Autonomous navigation in environments where conditions are constantly changing is restricted to very low speeds,” explains Matthias Müller, Lead of Embodied AI Lab at Intel Labs. “This makes drones unable to operate efficiently in real-world situations where something unexpected may block their path and time matters.”That’s obviously a big impediment to safely rolling out drones for commercial use. The solution seems to be harnessing the decision-making abilities of expert pilots to train drones to function autonomously.”In partnership with the University of Zurich, we were able to show how a drone trained exclusively in simulation by imitating an expert pilot is able to perform in challenging real-world scenarios and environments that weren’t used during the training of the convolutional network,” says Müller. “The trained autonomous drone was able to fly through previously unseen environments, such as forests, buildings and trains, keeping speeds up to 40 km/h, without crashing into trees, walls or any other obstacles – all while relying only on its onboard cameras and computation.”The results were achieved by having the drone’s neural network learn from a simulated expert pilot that flew a virtual drone through a simulated environment full of complex obstacles. The expert had access to the full 3D environment while the drone’s neural network only had access to the camera observations with realistic sensor noise and imperfect state estimation. That input imbalance (what researchers call a “privileged expert”) forced the drone to learn to act with exceptional dexterity in less than ideal conditions. The quadrotor demonstrated a decrease in the latency between perception and action while simultaneously demonstrating resiliency in the face of perception artifacts, such as motion blur, missing data, and sensor noise.

[embedded content]

“Existing systems use sensor data to create a map of the environment and then plan trajectories within the map – these steps require time and errors compound, making it impossible for the drones to fly at high speeds,” says Müller. “Unlike current systems, future drones could learn navigation end-to-end in a simulated environment before going out into the real-world. This research shows significant promise in deploying these new systems in a wide array of scenarios including disaster relief, construction sites, search and rescue, agriculture and logistics, and more.”

One of the benefits of this system is its applicability to a wide variety of real-world environments. The approach demonstrated in the research involved experiments that tested in a set of human-made environments (e.g. simulated disaster zone and urban city streets) and also diverse natural environments (forests of different types and densities and steep snowy mountain terrains). Future application areas for the technology could be disaster relief scenarios, construction sites, search and rescue, agriculture and logistics, and delivery. More

Matt Rutherford was just a few days into his planned unaided, non-stop solo-sailing trip around North and South America when he realized he’d left all his extra pants on the dock. The days of preparation before setting off had been frantic and a few things got left behind. This was a problem. He was facing 309 days at sea with little human contact and his small 27-foot sailboat, which he got for free and outfitted himself, was designed for bay sailing and not the notoriously unrelenting weather and towering seas of Cape Horn or the perilous ice of the Northwest Passage. To cap it off, he had just spilled diesel fuel all over himself, the result of a leaking fuel bladder, and he really wanted a change of clothes. Most people would have turned back. Rutherford, whose blue collar background and gruff presentation put him at odds with the Instagram sailing culture that’s proliferated among a new generation of seafarers, grumbled something under his breath, checked his course, and pressed ahead on a trip that would secure his place in the record book.That 2011 trip, during which Rutherford gained enormous respect for polar regions and for the surprising utility of small sailboats, would catalyze a mission that may seem odd in a digital world: To use sail power for science, and in so doing to prove that sailboats, one of the original technologies that helped humanity expand its horizons, are ideal platforms for next generation data collection in the world’s oceans.When I did the trip around Americas, the ocean became my home. I was no longer visiting the ocean, I was living at Sea. When I got back I wanted to do something that gives back, something that allows me to grow and utilize my skills as a sailor. I wanted something that I could work on for the rest of my life.Rutherford, a one-time drug dealer (he’s open about that fact in interviews) who watched his childhood friends get locked up or killed one by one before deciding life had different plans for him, does not have a science background, but on his adventure around the Americas he came face to face with the front lines of climate change. “Ground zero for the observations of climate change are in the polar regions,” he says. “That’s where you see the glaciers retreating and the ice melting.” Combined with frequent run-ins with garbage, it was an eye-opening experience.He also confronted a reality that has stymied scientists for generations: These places are phenomenally difficult to traverse, particularly via boat. “You know, it’s uncharted, you don’t know where the rocks are and there are rocks all over the place, compasses don’t work, the weather forecasts are garbage, there’s ice, there’s fog, there’s polar bears that will eat your head. So it’s a really unique challenge.”Those factors contribute to the astronomical price tag of doing science at sea in remote regions. The average cost of a scientific research vessel is estimated at $25,000 per day. In the arctic, that price tag can double, easily reaching or exceeding $50,000 per day. Almost without exception scientific expeditions are done on power vessels, and fuel accounts for a substantial portion of the total costs. It also limits the effective range and duration of expeditions, which must be able to safely return to shore to refuel with relative consistency. That significantly limits the kinds of data collection scientists are able to do at sea.  In 2013, Rutherford teamed up with scientist Nicole Trenholm to found Ocean Research Project, a 501(c)3 dedicated to scientific exploration under sail. We are forging a new path of discovery by combining modern technology with age-old efficiency to create lower cost research expeditions that provide an effective alternative to the more costly, big science paradigm. ORP’s research objectives are born from close collaboration with scientists focusing on the most important issues.

Sail power is having something of a renaissance. Mounting pressure to reduce carbon emissions in shipping have led to renewed interest in wind-assisted ships. Fixed-sail propulsion designs have been proposed for a variety of large ship applications. Autonomous data collection platforms, including underwater gliding drones and sail drones, are making their way into the modern naturalist’s toolkit. One problem with these drones, interestingly, is that though they’re small and easy to launch, it’s expensive for scientists to mount expeditions for launch and recovery, events that can be weeks apart.Ocean Research Project (ORP) is born from the same spirit. The organization designs unique expeditions that enable small, core teams of researchers to gather critical information from remote, sensitive areas worldwide. During an early expedition in 2013, Rutherford and Trennholm spent 70 days in the Atlantic to survey the eastern side of the North Atlantic Garbage Patch, which at that time was unmapped. “One of the reasons it hadn’t been mapped, and why we were out there so long, is you have to sail all the way to basically the Azores before you can even start,” says Rutherford. As any recreational sailor knows, that kind of long range expedition is perfectly suited to a sailboat, which moves slowly but inexpensively and can sustain a small crew essentially indefinitely with the right kind of equipment and food supplies.Ocean Research Project is decidedly a bootstrapped affair, in accordance with the lifestyle that has taken Rutherford around the world. It is grant and donation funded and has relied on boats that are begged, borrowed, or bartered. These are often small boats originally designed for short hops in protected waters. Rutherford and Trenholm strip them down and remake them into purpose-built research sailboats. In January 2019, Rutherford was taping an episode of his podcast, Single-handed Sailing, which is a stream-of-consciousness affair that has a cult following among a certain kind go-it-alone sailor, when he began describing his ideal boat for ORP expeditions. After the show he got an email from a listener who knew of someone with a 65-foot steel sailboat, exactly the kind Rutherford had described. The boat was a home-build project and Rutherford was understandably wary, but when he got there he was delighted. The boat had masts installed and seemed 70% of the way there. ORP has been outfitting and rebuilding it ever since — an arduous process funded by word-of-mouth donors and support of Rutherford’s podcast.”We need to have the boat in the water early April and then we got to leave for Greenland probably by mid-may and we have about a half million dollars of scientific equipment this time.”That includes a multi-beam sonar for mapping the ocean floor in accordance with the UN Seabed 2030 Project, as well as equipment to measure glacial sediment and resulting nutrient blooms in arctic waters. ORP will also be ground truthing satellites for NASA, conducting microplastics research, and doing a variety of water sampling.The estimated daily cost of the expedition will be $3000 per 24 hours, a small fraction of a traditional research vessel.Interestingly, Rutherford sees a blending of the newest and oldest technologies as an ideal approach to data collection. Aerial, aquatic, and submersible drones, for example, are a perfect use case for his organization’s sail expeditions.”Fully autonomous data collecting robots are going to play a large role in the future of ocean research and they’re already playing a bigger bigger role every single year. These Technologies are not massive and are perfectly supported on, say, a 65-foot boat at a fraction of the cost. And really that’s where you show the true capability of the sailboat being a professional data collection platform. It’s slower in important ways for probes and mapping the seafloor, it’s much more cost-efficient, and it’s the integration of these these new technologies that are really going to take off as time goes on.”Visit the Ocean Research Project website for information on how you can support the organization’s work. More

Indy Autonomous Challenge

Innovation

No matter which team places first at the upcoming Indy Autonomous Challenge (IAC), the real winner will be the open source architecture powering the cars. The autonomous racing event, which takes place this week, pits nine teams representing 21 universities against each other in high-stakes racing for a $1 million prize purse.”To our knowledge, all of the vehicles in the IAC are running ROS 2 and Autoware as the basis of their autonomy stack,” says Katherine Scott, Developer Advocate at Open Robotics, which oversees the Robotic Operating System (ROS), an open source architecture for robotics development. Autoware is the first all-in-one open source software for self-driving cars.Organized by Energy Systems Network and the Indianapolis Motor Speedway, the primary goal of the IAC is to advance technology that can speed the commercialization of fully autonomous vehicles and deployments of advanced driver-assistance systems (ADAS). Much like racing development often leads to innovations adapted for the street, the high speeds and incredible handling challenges of racing are ideal proving grounds for autonomous mobility. Indianapolis-based Lilly Endowment Inc., one of the world’s largest philanthropic foundations, provided a grant to help Energy Systems Network develop the IAC and fund the $1 million IAC prize purse.

[embedded content]

But the unsung hero of the event is the open source architecture that’s made it possible, in relatively short order, for teams of university students to develop autonomous controls paradigms for performance race cars. “What we’re really excited about in the IAC — more so than just the high-speed vehicle autonomy — is the prospect of having an entire generation of engineers learning and using ROS,” says Scott. “Advanced technology doesn’t happen in a vacuum; students need a common language and set of tools to work together efficiently. When done right, open-source technology is accretive, and the advances in this year’s competition will likely make it into subsequent competitions and perhaps production vehicles.”The car used for the event, which is an incredible feat of racing engineering, is the Dallara-produced AV-21 that has been retrofitted with hardware and controls to enable automation. Since 2001, Dallara has been the sole supplier of the Indy Lights series.The performance race cars are equipped with custom sensing packages. Thanks to ROS and Autoware, the development process has focused on higher-level controls work.

“The combination of ROS 2 and Autoware have given the IAC student teams a firm starting point for building out vehicle autonomy. Rather than having to focus on building interfaces to the sensors and actuators of the vehicles, the students can focus on the more challenging control and planning tasks. For example, instead of spending their time building an interface to the vehicle’s LIDAR, the students can instead focus on using data from that LIDAR to keep the vehicle in its lane and avoid obstacles.”The competition takes place on October 23 and is open to fans via a registration page. More

Laevo Exoskeletons

Innovation

A Dutch exoskeleton company is launching an updated model of an innovative unpowered exoskeleton for workers. The new product from Laevo, called the Flex V3, is compelling for a couple of reasons, but the big one may be that it’s attempting to solve well-known problems that have kept exoskeletons, once vaunted as a breakthrough technology, a primarily niche product.The Laevo device is a passive, unobtrusive exoskeleton designed to alleviate strain and adapt to every posture. By leaning against the chest pad, the Laevo transfers force from the rest of the body to the thighs.In general, exoskeletons have stumbled from a business perspective. Powered and unpowered suits designed to augment human strength or reduce the strain associated with bending and lifting have been around since the early 2010s. One of the visions of robotic exoskeletons, which are marvels of biomechanics and engineering and were pioneered in robotics hubs like the Human Engineering Lab at UC Berkeley and companies like Ekso Bionics, was to replace wheelchairs for those with stroke or spinal cord injuries. But the market case for mobility applications has never quite worked, in large part due to the high price tag of robotic exoskeletons compared with the substantially lower price point of wheelchairs, a proven and resilient technology.More recently, exoskeletons have been marketed for work. Ekso Bionics, for example, debuted an upper-body exoskeleton at a Ford factory in 2018, and the technology has seen use in heavy industries and shipyards.

[embedded content]

That’s held out hope among developers of a commercial use case, and in that conception powered exoskeletons could be a bridge between a completely human workforce and the eventual arrival of the lights out, fully automated factory.Laevo, which sold a previous version of its unpowered exoskeleton for around $2500, is targeting different types of users, including those in non-industrial settings, such as logistics warehouses where workers move relatively small packages but are subject to repetitive stress strains. The lower price point of unpowered exoskeletons versus their robotic counterparts significantly lowers the threshold for adoption. Given that the logistics sector is increasingly coming under scrutiny in the U.S. for labor practices and working conditions, the timing could be right for a device designed to ease the physical strain.In the next months, Laevo will start rolling out the Laevo Flex in several markets across Asia, Europe, Latin America, Middle East and the U.S. The company says it already has about 2,500 units in the market. More

Masten

Innovation

Landing a spacecraft on the lunar surface is tricky. As any engineer can tell you, dust doesn’t help.A regolith, billions of grains of crushed rock are kicked up during a rocket-assisted landing, a huge impediment to creating reliable, reusable landers. Dust is also a notorious killer of robots, which is a critical issue for the future of space exploration and commercialization. Compounding the problem, commercial missions currently in the conceptual phase are more sophisticated than previous lunar missions, and the size of the lunar landers and the power of their engines has grown. That’s only going to make the problem of space dust, which can rip through the sheathing and clog engines and instruments, that much worse.One solution is to build landing pads on the moon, but that’s incredibly expensive ($120 million per pad, according to one estimate). An alternative may be to use a lander’s descent to create a landing pad in real-time.That’s just what a company called Masten Space Systems, which we’ve been tracking on ZDNet, is testing. The idea is to create near-instantaneous landing pads by injecting ceramic particles into rocket plumes to form a coating over lunar regolith as a lander descends on the lunar surface. It’s a bit like giving the lander a Midas touch — only a ceramic touch — and it could significantly reduce dust issues for future missions.The presto-landing pad concept is called the in-Flight Alumina Spray Technique (FAST), and the company has been testing it for several months with positive results.”Following our Phase 1 NASA Innovative Advanced Concepts award, we’ve spent the last year studying and advancing the FAST concept in collaboration with Honeybee Robotics, Texas A&M University, and the University of Central Florida,” according to a recent company post. “And we just wrapped up our initial research, proving the solution is feasible in the lunar environment.”Masten has been busy thinking up innovative concepts to commercialize lunar travel. The company recently unveiled its GPS concept for the moon, and it’s working on ways to extract water using lander engines — ironically, a process that would intentionally displace significant volumes of regolith.

For the current test, Masten figured out the thickness required for the pads to work effectively.As an example, a large-scale Artemis human landing system would require alumina particles of approximately 0.5 millimeters diameter to pass through the engine without melting. The particles would impact the lunar surface at approximately 1,500 meters per second to create an initial base layer on the lunar surface that’s approximately 1 millimeter thick.After the base layer is deposited, alumina particles of approximately 0.024 millimeters in diameter would be required to heat up and liquify as they pass through the engine. These particles would impact the surface at approximately 650 meters per second and create additional layers that build up and strengthen the landing pad. The full deployment would take 10 seconds to release 186 kilograms of alumina at up to 30 meters above the lunar surface, creating a 6-meter diameter landing pad. The pad would then require 2.5 seconds to cool before the vehicle touches down for a safe landing.If that kind of calculation doesn’t make the engineering problem solver in you light up, it’s time for a vacation and a recharge. Space is once again home to some of the most creative engineering and applied chemistry in the world, thanks to the burgeoning commercial space sector. The solution is still a long way from implementation, but Masten has big plans for the moon and beyond. The company is also ambitiously turning its sights on Mars.”In Phase I, we advanced the technology readiness and laid the groundwork for future development. Our goal is to further mature the landing pad technology by testing it in a lunar environment in the next phase. Looking even further ahead, the FAST concept can be applied to other planetary bodies like Mars where loose regolith also poses risks to human and robotic missions.” More