Robotics

Getty Images/iStockphoto
Quality in manufacturing is mission critical. AI-powered quality inspection is nothing new, but a joint venture from two big players in manufacturing could markedly improve outcomes and reduce barriers to entry.The new venture is called Lean AI. The technological secret sauce is what’s known as unsupervised AI, which is a cutting edge deep learning technology that doesn’t require massive datasets, months of setup time, or known inspection paradigms to function. The new company is a collaboration between Johnson Electric, which has knowledge and experience in manufacturing, and Cortica Group, which has pioneered unique autonomous AI technology for visual inspection.”With the power of Cortica’s Autonomous AI technology, and JE’s vast knowledge of the market, Lean AI will deliver a product that reduces the cost of human error when it comes to quality inspection in manufacturing and address the vulnerabilities in the current market,” says Karina Odinaev, CEO of Lean AI.The problem is that conventional Deep Learning-Based Quality Assurance Systems can take weeks or months, to deploy and rely on a data scientist or AI experts feeding large manually tagged training sets with thousands of defect image examples. These systems require constant maintenance and re-training for product variations or new cameras.Lean AI is leveraging a newer generation of unsupervised deep learning-based quality assurance technology to get past existing challenges. Its unsupervised system uses unlabeled data, applies predictive quality assurance, and compiles data that increases the speed of deployment and scaling. It’s an open platform, meaning it’s agnostic to camera, defect type, and product. That flexibility marks a big evolution in AI-driven inspection, which is a massive and growing market, particularly with renewed emphasis on efficiency as supply chains are stretched thin.By some estimates, the global machine vision market is currently valued at US$11 billion and is forecast to increase to US$15.5 billion by 2026.”Cortica has developed self-learning AI that is fundamentally different from traditional deep learning systems. Autonomous AI Technology operates like a human brain — it’s not a fixed system; instead, it continuously adapts itself to various scenarios and learns online in real-time. Its technology requires far less computing power, can be deployed at a fraction of the cost, and provides far superior performance outcomes,” says Igal Raichelgauz, Founder and Chairman of Cortica. “Our technology is robust and generic and applicable within a multitude of signal domains such as visual, audio, time series and other domains; visual inspection is only the beginning. Autonomous AI technology is quickly becoming the benchmark for the industry.” More

Brain Corp
The robot takeover won’t happen like we imagined. In fact it’s already well underway, albeit in sectors that are likely peripheral for most ZDNet readers. (Don’t worry … our time will come.)Case in point, over the last year robots have replaced workers at notable scale in the commercial cleaning industry. A company called Brain Corp, which creates core technology in robotics, recently announced that from October 1, 2020, to October 1, 2021, usage of automated robotics increased dramatically across a number of industry sectors, particularly in hospitals (+2,500%) and education (+426%).The news comes in the middle of the ‘Great Resignation’ within the U.S. and across the world, which has impacted the ability for those across industries to employ and retain workers. The U.S. Bureau of Labor Statistics stated that 4 million Americans quit their jobs in July 2021 and a study detailed in the Harvard Business Review found that, in general, resignation rates were higher among employees who worked in fields that had experienced extreme increases in demand due to the pandemic. Shifting office cultures and changing demand during the pandemic has had an especially large impact on janitorial services.Enter the robots. The table was set for a technology-driven overhaul. Autonomous mobile robots (AMRs), which can map and navigate semi-structured environments with ease, are now prolific in industries like logistics, where they ferry packages around warehouses, and in grocery stores, where they scan shelves to give retailers data driven insights on merchandise. Cleaning robots have been around for some time, and companies like Brain Corp have been refining the technology via AI software that allow autonomous robots to operate in new, varying environments, as well as advanced sensors, which have fallen in price as AMRs proliferate in the market. It’s a classic automation adoption scenario, one that combines drastically shifting labor economics, the pressures of a global pandemic, and the opportunity afforded by technological development. In the case of Brain Corp, the company’s BrainOS-powered fleet of autonomous mobile robots just reached 100 billion square feet of coverage a significant milestone that the company estimates is equivalent to 6.8 million hours of human work.”We are thrilled to celebrate reaching 100 billion square feet of coverage with our fleet which represents the square footage of the entire commercial space in the United States,” says Eugene Izhikevich, CEO and Founder of Brain Corp. “The milestone represents a clear success of deploying autonomous robots at scale and across multiple industries. As a company committed to continuous improvement, it’s been incredibly gratifying to see such major advances in our fleet’s performance, even as it has been scaled and expanded to operate in multiple new dynamic public environments across the world.” The same mechanisms driving automation in commercial cleaning are pushing development in spaces like fast food, where burger robot Flippy is becoming a viable cooking option for fast food chains, and window washing, where autonomous robotic arms can perform one of New York City’s most iconic jobs.

It’s a sign that the robot revolution isn’t coming. It’s already here. More

Starship Technologies

Innovation

Another campus, another rollout of roving delivery robots. You may not know it, but delivery robot vendors are making a play for campuses across the country in a bid to grab a market toe-hold in relatively structured environments free of much of the regulatory complications of municipalities.Starship Technologies has delivered 30 autonomous robots for food service to South Dakota State University in the latest example. The robots will deliver from three campus vendors — Grille Works, Papa Johns, and Starbucks — with additional locations added soon.”The one thing we have learned in recent years is that students and faculty like flexibility in their dining options,” said Doug Wermedal, associate vice president for student affairs at SDSU. “The ability to have something delivered to various locations throughout campus and the community will be impactful to our students and employees as they continue to manage busy and demanding schedules. We are excited about this partnership, the robotics technology and the student employment opportunities Starship will bring to our campus.”But does a campus of 14,000 students and faculty need 30 delivery robots? With ongoing concerns about clustered dining during the pandemic, there’s some case to be made for the flexibility and public health benefits of contactless delivery. One thing is for certain, and that’s that Starship Technologies has identified college campuses as important strategic markets in a highly competitive delivery paradigm and shifting regulatory considerations.Starship already has robots on the campuses of Arizona State University, Purdue University, George Mason University, and Northern Arizona University. Since its launch, all campuses have increased the number of robots, dining options, and hours of operation to meet the high demand for the service.While the number of robots deployed on campuses isn’t a show stopper, the value to the company is exceptional. In many ways, colleges are the perfect test bed for delivery robots. Students tend to live well within a 30-minute delivery radius. Integration with meal plans, which is the model governing the SDSU rollout, helps ensure a ready customer base, and participating universities are easily wooed by the allure of being a forward-thinking institution with Silicon Valley connections (Starship is headquartered in the Bay Area). Campuses also offer an exceptional proof of concept for a variety of Starship’s constituents, from investors to prospective customers to regional regulatory bodies that are approaching robot delivery with appropriate caution. Halfway through 2021, Starship announced that it had repeatedly set delivery records in its campus deployments during the pandemic.

“I hadn’t even heard of robot delivery before I started school, and now I don’t see a future without it,” said Claire Sunderman, a student at Bowling Green State University, where the company has a deployment. “I’d be perfectly happy to have a robot deliver a lot more things because it would save me so much time. Now that I am graduating, I will really miss the convenience and seeing the robots on campus everyday — I wish I could take one with me!”  That sort of attitude bodes well for companies like Starship, which aren’t just proofing technology but also training a new generation of adopters. Other autonomous delivery companies have adopted similar tacks. A company called Flytrex, for example, made headlines by offering food delivery via drone at a golf course in North Dakota recently. The access-limited space permits management to collect waivers from golfers, which allows Flytrex to avoid strict FAA regulations when operating over public areas.But ultimately, these testbeds, while good for short term adoption and product refinement, aren’t sufficient to sustain these companies. For widespread adoption to occur, automation firms still need to tackle the thorny issue of local regulatory hurdles. So far, companies like Starship (either out of prudence or because they don’t have deep enough pockets) have avoided the blitz mentality of Uber and Bird, which left local regulators scrambling to react and have opted instead for a more methodical rollout. More

Unlimited Tomorrow

Innovation

Here’s the problem: Advanced prosthetics can improve the quality of life for amputees, but the devices are incredibly expensive. A company called Unlimited Tomorrow has created a process to solve for this, and in so doing, it’s become an important test case for technology-driven manufacturing.The problem isn’t insignificant. Nearly 2,000,000 people live with limb loss in the United States, which records approximately 185,000 amputations each year, alone, per the Amputee Coalition.Historically, amputees have been hard-pressed to find lightweight and comfortable prosthetics, intuitive and reliable, and, perhaps most importantly, affordable. These challenges are compounded for children with limb loss, as they’re required to purchase multiple prosthetics as they grow. It costs an average of $80,000 per limb to keep a child outfitted with an appropriate prosthetic.To make prosthetics that are tailored to individual amputees’ needs, and particularly children, at a fraction of the cost, Unlimited Tomorrow Founder Easton LaChapelle has turned to emerging technologies, such as Siemens’ 3D printing software. It’s a game-changer.

[embedded content]

“Until recently, all bionic prosthetic devices were produced using expensive tooling, molds, and materials,” LaChapelle tells me. “This type of production limits the sizes and form factors that can be offered and adds cost. By nature, a prosthesis must be durable and robust because it is used daily. This has made it difficult to develop alternatives until 3D printing became an option. Other technological barriers include batteries and sensors used to control the prosthesis.”Interestingly, the story of how bionic devices are typically made is a story of the inefficiencies of manufacturing segmentation.”We have found that a lot of the costs come from segmented manufacturers that all create different components that go into a device,” explains LaChapelle. “Each of these manufacturers needs to make a margin, and overall, this increases the costs to the clinicians. Clinicians are critical to creating a sound prosthetic device. They work directly with the patients, create the socket (the most critical part of the device), and fabricate and assemble the final unit. These clinicians then need to bill the insurance companies to make their margin. When you add all of this together, you get expensive prosthetic devices.”

Therefore technological improvements will help reduce these costs, but technology alone won’t solve all the issues, a point that’s often left out of glowing predictions of technology-driven manufacturing efficiencies. Unlimited Tomorrow identified cost structure as an issue early on and solved these challenges by developing its technology and manufacturing from the ground up. This allows the company to control costs very accurately. Unlimited Tomorrow has an in-house clinical team, as well as production and manufacturing teams. Having everything under a single roof allows the company to be very efficient, control quality across the board, and reduce costs tremendously compared to other companies and services. By solving for segmentation and wholly owning its manufacturing, Unlimited Tomorrow is solving for the segmentation-induced costs. And of course, the very idea of a small enterprise owning its own manufacturing is a product of technology, part of the virtuous circle of technology enabling processes that drive new utilizations of technology.Says LaChapelle, “3D printing, 3D scanning, and software have enabled us to produce prosthetic devices that are unique to each person (in terms of shape, size, and skin tone), extremely durable, lightweight, and affordable. The devices we make have comparable functionality to the most advanced devices on the market. Through a convergence of these technologies, we can offer them at the lowest cost within the bionic category.”Technology has also contributed to astounding customization. For example, unlimited Tomorrow offers 144 skin tones, and the devices are custom-made in a mirror image of the end user’s opposing limb. 3D scanning is used to capture the user’s residual limb geometry, and new software and algorithms allow the company to create sockets instead of much costlier hand fabrication digitally.Siemens, which put Unlimited Tomorrow on my radar, is playing a role in this manufacturing success story. “Siemens’ software is critical in our process. Their software takes in the raw 3D scan from our users and allows our clinical team and socket generation team to digitally fabricate and fine-tune sockets for a wide range of users. No two residual limbs are ever the same, and to have software that can account for that, while also making it easy enough for our team to produce sockets, is incredible.”Unlimited Tomorrow has created an upgrade program for children whose devices will become obsolete as they grow. “We upcycle the quality and expensive components and create a new device for them at half the cost.”It’s another example of the flexibility that an integrated manufacturing process brings, and it’s a good illustration of how one company is able to serve such a variety of customers. “Our youngest user is seven years old — our oldest is 86 years old!” exclaims LaChapelle enthusiastically. More

Fabric

Digital transformation

Pioneering “micro-fulfillment” company Fabric, which has been something of a darling among investors, just passed an important milestone with a valuation that exceeds a billion dollars. The company uses robots and geographically strategic fulfillment centers to get products to customers in a hurry, as quickly as one-hour after an e-commerce order is made.Is the company the future of fulfillment in a world that previously has been dominated by Amazon? Investors certainly seem to think so. The company has netted a whopping $336 million in total funding on the promise of democratizing fast fulfillment.The reason for the investor enthusiasm is readily apparent. This year saw e-commerce sales penetration more than double to 35%, fueled by the COVID-19 pandemic’s acceleration of existing online shopping trends. The same-day delivery market in the US is poised to grow by $9.82 billion over the next four years. Meanwhile, supply chain catastrophes and strained fulfillment capacity has created a bottleneck. Fabric thinks it can solve the problem with a fully integrated micro-fulfillment process powered by AI and robots.”At the center of this perfect storm of e-commerce is Fabric and our ability to enable on-demand retail at profitable unit economics,” says Elram Goren, Fabric CEO and co-founder. “We see this milestone as a real turning point in the industry, from what was once intrepid exploration of micro-fulfillment to total market validation and now rapid expansion.”The recipe to democratize the last-mile logistics sector, according to Fabric, is a blend of high tech fulfillment robots and smaller-than-average fulfillment centers located in urban zones close to customers. The idea is that the physical remoteness of typical logistics facilities prevents most retailers from offering true on-demand delivery outside select major metropolitan markets. But by harnessing networks of tiny automated hubs, micro-fulfillment could enable retailers to store their goods in the hearts of cities while still benefiting from the efficiency of automation.A couple of years ago, Fabric (then CommonSense Robotics) proved the concept with its first 1-hour fulfillment delivery, which is made in partnership with Super-Pharm, an Israeli health and beauty retailer. Fabric runs micro-fulfillment operations for grocery and general merchandise retailers in New York City, Washington, DC, and Tel Aviv. The company recently announced major partnerships with Walmart, Instacart, and FreshDirect. It’s no surprise that all of those companies are competitors of the many-headed Bezos hydra. There seems to be a fully dawned recognition within the retail space that Amazon’s shrewd move to corner the logistics market can’t be replicated or challenged by any one retailer. Smaller enterprises have a unique opportunity to create their own logistics operations and compete on customization and customer experience, which is what startup PetFriendly has done. But players like Walmart, which aren’t capable of competing on customization, need massive fulfillment technology infrastructure, and it’s far more attractive to team up with a technology-focused service provider than it is to create that infrastructure from scratch, a risky bet.

Investors have responded favorably to Fabric’s position in the marketplace. The latest $200 million Series C funding round was led by existing investor Temasek, with participation from Koch Disruptive Technologies, Union Tech Ventures, Harel Insurance & Finance, Pontifax Global Food and Agriculture Technology Fund (Pontifax AgTech), Canada Pension Plan Investment Board (CPP Investments), KSH Capital, Princeville Capital, Wharton Equity Ventures, and others.”We believe the movement to local fulfillment presents an opportunity to make retail and e-commerce more sustainable, and we’re thrilled to partner with the leader in micro-fulfillment to make this vision a reality,” says Eric Kosmowski, Managing Partner at the Princeville Climate Technology Fund. “By leveraging existing real estate with a small footprint in close proximity to end consumers, utilizing more sustainable packing materials, and minimizing shrink and waste through smart inventory management, Fabric’s micro-fulfillment centers could lower last-mile emissions significantly.” More

Miso Robotics

Innovation

A new version of the robotic fry cook looks like the odds on favorite to automate fast food production. Miso Robotics, the creator of Flippy, the automated fry cook that hangs from a rail system over a conventional griddle, just announced a new version of its flagship robot, dubbed Flippy 2.With a pressing labor shortage and booming drive-thru and carryout demand over nearly two years of pandemic-induced recalibration, fast food has increasingly turned to technology to increase efficiencies and improve consistency and output. McDonald’s has been a leader in this arena, and just this week announced it was teaming up with IBM to create an automated drive-thru concept. Miso Robotics, an unusual success story in the crowd-funded technology space, is positioning itself as the go-to provider for automated cooking. The concept makes a lot of sense: fast food is built on repeatable, high-output cooking processes that increasingly rely on complex order management and prediction capabilities to reduce wait times and spoilage. Labor markets have been unpredictable over the past couple of years, and many fast food companies feel incentivized to look for a technological solution when faced with increasing minimum wages.The evolution of Flippy, the robotic fry cook, designed to automate fast food production, has been fun to watch. First, the robot, which looks decidedly like something out of The Jetsons, did a handstand, going from the floor to an overhead rail system in response to space constraints in the kitchen. There were key updates to its AI and an expanding menu of foods it could prepare to crispy perfection. Along the way, the robot got high marks behind the grill at some iconic chains, like White Castle, the original burger chain. Just recently, Miso Robotics, Flippy’s creators, spun out a chicken wings-specific model to service partners like Buffalo Wild Wings.Flippy 2 includes improvements that largely came from the company’s feedback from White Castle, with which Miso Robotics has been partnering for about a year.”We learned so much from Flippy and our partnership with Miso Robotics. It’s amazing to see the future of how we provide even better service and even more hot and tasty food to our craving customers happening right before us in real-time,” says Jamie Richardson, Vice President at White Castle. One improvement is the new AutoBin system for lower volume and speciality foods like onion rings or chicken tenders, providing restaurants with a more capable and complete frying solution. Each bin can hold as much as a full fry basket, be customized for a kitchen’s specific needs, and be delineated for individual products like vegetables and fish to prevent cross-contamination.

Flippy 2 also features a new design that takes up less space within the kitchen, which the company says accounts for 56% reduced aisle intrusion, 13% height reduction and fewer overall cleanable surfaces.”Like all technologies, Flippy 2 has evolved significantly from its predecessor, and we are extremely grateful for the insights collected from White Castle to truly push its development forward in a real restaurant environment,” says Mike Bell, CEO of Miso Robotics. “Flippy 2 takes up less space in the kitchen and increases production exponentially with its new basket filling, emptying and returning capabilities. Since Flippy’s inception, our goal has always been to provide a customizable solution that can function harmoniously with any kitchen and without disruption. Flippy 2 has more than 120 configurations built into its technology and is the only robotic fry station currently being produced at scale.”     Miso Robotics has several other pilot agreements with leading national brands in place, including its recently announced partnership with Inspire Brands. The company is hoping its technology will lead the way in a technology shift already happening in the $278.6 billion market. More

UBTech

Artificial Intelligence

Robots had a big year. Amid uncertainties related to the pandemic, chronic labor shortages, supply chain catastrophes, and renewed emphasis on touchless and autonomous service, the adoption of robotic systems has steadily increased.Supply chains are rapidly digitizing amid Industry 4.0 transitions. Competition and collaboration among rivals, particularly in the industrial robotics space, drives innovation and lowers the adoption threshold for automation solutions. Robots are also entering new markets, including food and beverage, textiles, wood products and plastics. Companies like Sarcos are transforming construction and infrastructure, bringing new efficiencies to old sectors, and the logistics space, which increasingly relies on autonomous picking and sorting solutions and mobile ground robots, is seeing explosive growth as habits shift further toward e-commerce.”As we come out of a global pandemic and continue to experience more issues with global supply chains and labor shortages like we see now,” John Rhee, SVP and General Manager of UBTECH North America, tells me, “the adoption of robotics will grow at a rate faster than previously forecasted.”I connected with Rhee, whose company makes robots for a variety of uses, including disinfection and education, to get his insights on how the past year shaped the automation sector, as well as what’s in store in the year ahead.
ABB
Robotics in manufacturingWhat were the most important milestones for robotics in manufacturing in 2021?John Rhee: For UBTECH, we were able to utilize our company’s strengths in both robotics as well as vertically integrated manufacturing in order to create solutions for the global pandemic in 2021. We recognized the ongoing pain points in 2020 with the pandemic by developing robots that performed tasks like surface cleaning and temperature monitoring in high-risk environments. That allowed us to identify a bigger need and move into robotics using UV-C light to provide a more cost-effective and less toxic method to sanitize autonomously. UBTECH was able to create a disinfection solution that is one of the highest intensity mobile options in the market today AND for the lowest cost on the market. Based on the current and future needs we were seeing and experiencing in the global market, we looked to leverage UV-C technology that hospitals and government municipalities have been using for decades. In creating the ADIBOT disinfection system, our goal was to provide this option to a wider group of organizations like school districts and businesses to help in their efforts to keep children and staff protected against COVID-19 c.diff, MRSA and more.Object manipulation

How has object manipulation advanced over the past couple of years? What role will AI play, and what will robots be able to do soon that they can’t do now?John Rhee: With our Walker product, we’ve been able to show over the past few years how object manipulation has gotten more advanced within robotics. Our first iterations of Walker from 2018 could kick a soccer ball, and now in 2021, our Walker X can hold groceries, carry objects, write, and play interactive games with different members of the household. The Walker X humanoid robot is the latest in these types of advancements that includes progress to actions like hand-eye coordination, multi-terrain manipulation, and the addition of multimodal emotional interaction. Additionally, we continue to develop bin picking technologies for various use cases across our robotics portfolio. All these advancements are driven by AI with more intelligent uses of computer vision and sensor systems. Robots will be more versatile and precise in both the work and home space, being able to perform more solid functions than in the past.
UBTech
Robots in the classroomA lot of parents might back at the idea of using AI in the classroom. Can you discuss the benefits and possible applications, as well as some things we should watch out for? John Rhee: AI is likely already in the classroom in some way, shape, or form for a lot of children today. An Alexa ask for gaming simulations or complex LMS systems that the teacher uses to manage the classroom and measure performance. We believe that students understanding what AI is and how it can be used to solve real-world problems is more critical than what a basic computer lab curriculum was for students in the 1980’s. In our education division, UBTECH Education, we want to ensure children are taught a full understanding around Artificial Intelligence that includes not only learning how to build robots while incorporating programming with AI but also how AI can be used in the real world in all circumstances. As a part of our core belief to prepare students for the jobs of tomorrow, our curriculum and products are in line with also equipping them with all the angles for the job at hand.  More

Augury
Augury, an industrial AI + IoT company—with customers such as Colgate, Hershey’s and Pepsi – has raised $180M in Series E funding. The round, led by Baker Hughes, launches the company to unicorn status and makes it an important bellwether for Industry 4.0 transitions.The deal comes amidst a pandemic-driven disruption in manufacturing and supply chains, which VCs have taken keen notice of. Something on the order of $45.1B has been raised by industrial start-ups so far this year, compared with the $34B raised in all of 2020, according to PitchBook. Augury’s wireless sensors and cloud platform monitors the “vitals” (like temperature and vibration) of critical industrial equipment. AI is used to compare them to a database of over 80K+ machine sounds so that faults can be detected before they cause downtime. It does this at an accuracy rate of over 99%. This is especially important as automation and robotics become driving forces of global manufacturing and distribution.”We’ve spent the last decade building towards a future where we can always rely on the machines that matter, in the sectors that matter,” said Saar Yoskovitz, co-founder and CEO of Augury. “Today marks a significant step into that future since our industry’s leading organizations have recognized the importance of Machine Health to them and their customers, and trust Augury to be their Machine Health partner. I’m thrilled by the opportunity this funding, coupled with the market access our new investors provide, gives us to further fuel Augury’s exponential growth and bring the impact of Machine Health to new markets.” In many ways the technology specializes in increasing efficiency by decreasing the costs associated with downtime and catastrophic equipment failure. Examples abound, but three touted by Augury include one in which Colgate-Palmolive saved 2.8 million tubes of toothpaste by avoiding a single machine failure; another in which ICL saved a million dollars in downtime and production loss costs at a single facility in less than 10 months; and one in which PepsiCo has rolled out the tech on all North American Frito Lay plants after Augury saved them 1M+ pounds of snacks.Baker Hughes and SE Ventures are investing in Augury at a time of astonishing disruption for the supply chains. Global manufacturing is dealing with unprecedented uncertainty, and in that environment Augury’s pitch of shoring up owned infrastructure is landing on friendly ears.According to the company, proceeds will contribute to Augury’s over $200 million war chest as it expands globally in sectors like energy while strengthening its core manufacturing customer base. More