Robotics & AI

Watch Google’s ping pong robot pull off a 340-hit rally • ZebethMedia

By zebethcontrolPosted on October 18, 2022Posted in Google, Ping Pong, Robotics & AINo Comments

As if it weren’t enough to have AI tanning humanity’s hide (figuratively for now) at every board game in existence, Google AI has got one working to destroy us all at ping pong as well. For now they emphasize it’s “cooperative” but at the rate these things improve, it will be taking on pros in no time. The project, called i-Sim2Real, isn’t just about ping pong but rather about building a robotic system that can work with and around fast-paced and relatively unpredictable human behavior. Ping pong, AKA table tennis, has the advantage of being pretty tightly constrained (as opposed to playing basketball or cricket) and balance of complexity and simplicity. “Sim2Real” is a way of describing an AI creation process in which a machine learning model is taught what to do in a virtual environment or simulation, then applies that knowledge in the real world. It’s necessary when it could take years of trial and error to arrive at a working model — doing it in a sim allows years of real-time training to happen in a few minutes or hours. But it’s not always possible to do something in a sim; for instance what if a robot needs to interact with a human? That’s not so easy to simulate, so you need real world data to start with. You end up with a chicken and egg problem: you don’t have the human data, because you’d need it to make the robot the human would interact with and generate that data in the first place. The Google researchers escaped this pitfall by starting simple and making a feedback loop: [i-Sim2Real] uses a simple model of human behavior as an approximate starting point and alternates between training in simulation and deploying in the real world. In each iteration, both the human behavior model and the policy are refined. It’s OK to start with a bad approximation of human behavior, because the robot is also only just beginning to learn. More real human data gets collected with every game, improving the accuracy and letting the AI learn more. The approach was successful enough that the team’s table tennis robot was able to carry out a 340-strong rally. Check it out: It’s also able to return the ball to different regions, granted not with mathematical precision exactly, but good enough it could begin to execute a strategy. The team also tried a different approach for a more goal-oriented behavior, like returning the ball to a very specific spot from a variety of positions. Again, this isn’t about creating the ultimate ping pong machine (though that is a likely consequence nevertheless) but finding ways to efficiently train with and for human interactions without making people repeat the same action thousands of times. You can learn more about the techniques the Google team employed in the summary video below:

DigestAI’s 19-year-old founder wants to make education addictive • ZebethMedia

By zebethcontrolPosted on October 18, 2022Posted in Battlefield, battlefield 2022, digestai, Disrupt, edtech, Robotics & AI, StartupsNo Comments

When Quddus Pativada was 14, he wished that he had an app that could summarize his textbooks for him. Just five years later, Pativada has been there and done that — earlier this year, he launched the AI-based app Kado, which turns photos, documents or PDFs into flash cards. Now, as the 19-year-old founder takes the stage for Startup Battlefield, he’s looking to take his company, DigestAI, beyond flashcards to create an AI dialogue assistant that we can all carry around on our phones. “If we make learning truly easy and accessible, it’s something you could do as soon as you open your phone,” Pativada told ZebethMedia. “We want to put a teacher in every single person’s phone for every topic in the world.” Quddus Pativada, founder at DigestAI pitches as part of ZebethMedia Startup Battlefield at ZebethMedia Disrupt in San Francisco on October 18, 2022. Image Credit: Haje Kamps / ZebethMedia The company’s AI is trained on data from the internet, but the algorithm is fine-tuned to recall specific use cases to make sure that its responses are accurate and not too thrown off by online chaos. “We train it on everything, but the actual use cases are called within silos. We’re calling it ‘federated learning,’ where it’s sort of siloed in and language models are operating on a use case basis,” Pativada said. “This is good because it avoids malicious use.” Pativada said that this kind of product would be different from smart assistants like Apple’s Siri or Amazon’s Alexa because the information it provides would be more personalized and detailed. So, for certain use cases, like asking for sources to use in an essay, the AI will pull from academic journals to make sure that the information is accurate and appropriate for a classroom. Despite running an educational AI startup, Pativada isn’t currently in school. He took a gap year before going to college to work on his startup, but as DigestAI took off, he decided to keep building instead of going back to school. Growing up, he taught himself to code because he loved video games, so he wanted to make his own — by age 10, he published a “Flappy Bird” clone on the App Store. Naturally, his technological ambitions matured a bit over time. Before founding DigestAI, Pativada built a COVID-19 contact tracing platform. At first, he just made the app as a tool for his classmates — but his work ended up being honored by the United Arab Emirates’ government. Image Credits: DigestAI So far, the outlook is good for the Dubai-based company. Pativada — who says he feels skittish about the CEO label, and prefers to think of himself as just a founder — has raised $600,000 so far from angel investors like Mark Cuban and Shaan Patel, who struck a deal on Shark Tank for his SAT prep company, Prep Expert. How does a 19-year-old in Dubai capture the attention of one of thee most well-known startup investors? A cold email. Mark, we apologize if this admission makes your inbox even more nightmarish. “I was watching a GQ video of Mark Cuban’s daily routine,” Pativada said. “He said he reads his emails every morning at 9 AM, and I looked at the time in Dallas, and it was about 9 AM. So I was like, maybe I should just shoot him an email and see what happens.” While he was at it, he reached out to Patel, whose educational startup has done over $20 million in sales. Patel hopped on a video call with the teenage founder, and by the next week, he and Cuban both offered to invest in DigestAI. “We raised our entire round through cold emails and Zoom,” Pativada told ZebethMedia. “It sort of helped because no one can see how young I look in person.” Before he decided to eschew college altogether, Pativada applied to Stanford and interviewed with an alumnus, as is standard in the admissions process. He didn’t end up getting into the competitive Palo Alto university, but his interviewer, who works at Stanford, did end up investing in his company. Go figure. “Our goal is to work with universities like Stanford,” Pativada said. The company is also targeting enterprise clients. Currently, DigestAI works with some U.S.-based universities, Bocconi University in Italy, a European law firm and other clients. At the law firm, DigestAI is testing a tool that allows associates to text a WhatsApp number to quickly brush up on legal terms. In the long term, DigestAI wants to create an SMS system where people can text the AI asking for help learning something — he wants information to be so accessible that it’s “addictive.” “That is what AI is — it’s almost the best version of a human being,” Pativada said.

Alaffia Health taps AI to detect errors in hospital bills • ZebethMedia

By zebethcontrolPosted on October 18, 2022Posted in Alaffia Health, Battlefield 200, Biotech & Health, Disrupt, Robotics & AI, TechCrunch Disrupt 2022No Comments

The multi-decade rise in healthcare costs isn’t expected to reverse course any time soon. In search of a fix, Adun Akanni and TJ Ademiluyi co-founded Alaffia Health in 2020, one of the startups participating in the ZebethMedia Disrupt Battlefield 200. The healthtech company uses machine learning to try to identify fraud, waste and abuse in healthcare claims. “We leveraged key insights from our family’s medical billing company in founding Alaffia,” Ademiluyi told ZebethMedia in an interview. “We determined that the majority of the waste in the system results from natural human error, lack of transparency in claims processing, and misaligned incentives between healthcare providers and payers. We founded Alaffia to tackle these issues using nascent machine learning and AI, built on top of deep healthcare domain expertise.” Alaffia sells services primarily to health insurance payers and enterprises that provide their employees health coverage. Using AI to extract and standardize data from hospital bills, including various medical billing procedure codes and dates of service, the platform aims to reduce payers’ spending by finding errors and overcharges within the bills sent by healthcare providers. The causes of medical billing errors are myriad, but often arise from double billing, missing the payer submission deadline and a failure to capture patient information. Non-specific diagnostic codes are another common issue, leading to instances of upcoding and undercoding. Upcoding is when a coder reports a higher-level service than patients received or never had performed, while undercoding is when billing codes don’t capture the full scope of work performed by a physician. Medical expenses are expected to grow by an average of 5.1% from 2021 to 2030, reaching $6.8 trillion, according to the Centers for Medicare and Medicaid Services — and a significant portion of those expenditures are derived from errors in health insurance claims. It’s estimated that about 80% of claims in the U.S. contain at least one medical billing error, and that as much as $300 billion is lost to provider fraud, waste and abuse each year. Image Credits: Alaffia Health “This is a quite challenging technical problem due to the lack of data standardization in the healthcare system, so we’ve rigorously trained machine learning models using training data generated by our in-house annotation team,” Ademiluyi said. Alaffia reviews facility bills for errors such as “unbundling” — i.e., using multiple codes for individual parts of a procedure — while checking the accuracy of more complex claims like implants and surgeries. The company says it taps registered nurses, certified coders and certified billers to cross-reference the AI’s findings, as well as a clinical review team that examines each claim and corresponding medical record. When asked about competitors, Ademiluyi says he sees “legacy industry participants” who manually process and review claims as Alaffia’s principal rivals. But Alaffia isn’t the only startup attempting to tackle the medical billing error problem with AI. Anomaly, which works with insurance companies and providers, offers an AI-driven platform designed to detect irregularities in medical bills. There’s also Nym, whose technology converts medical charts and electronic medical records from physician consultations into auditable billing codes automatically. Alaffia has managed to gain traction in the space, however — and funding. Ademiluyi claims the company’s services currently cover over 300,000 health plan members in aggregate. And to date, Alaffia has raised $6.6 million in venture capital from backers including Anthemis, 1984 Ventures, Aperture Venture Capital, Tau Ventures, Twine Ventures, Plug and Play Ventures and ERA’s Remarkable Ventures Fund. Ademiluyi says that 2022 revenue is on pace to more than double year-over-year. The near-term plan is to expand Alaffia’s commercial footprint and product offerings, he added, starting with hospital bill review services direct to patients. The company currently employs “just over” 20 people and expects to hire five more by the end of the year. “Fortunately, we operate in an industry resistant to recession. Regardless of pandemics, macro trends, or the outlook for interest rates, people will still visit the doctor to receive care,” Ademiluyi said. “When patients receive care, it leads to further healthcare spending, which benefits our business as we review generated hospital bills for errors. As we move into a slowdown in the market, large enterprises — both health insurance institutions and employers whom we support are actually looking at ways to lower their expenses, which we directly support by reducing healthcare spending. As such, we believe the pandemic and current slowdown in the economy to be a net positive for the business.”

Watch two Mini Cheetah robots square off on the soccer field • ZebethMedia

By zebethcontrolPosted on October 17, 2022Posted in Hardware, mini cheetah, robot, Robotics & AI, SoccerNo Comments

Some robotics challenges have immediately clear applications. Others are more focused on helping systems solve broader challenges. Teaching small robots to play soccer against one another fits firmly into the latter category. The authors of a new paper detailing the use of reinforcement learning to teach MIT’s Mini Cheetah robot to play goalie note, Soccer goalkeeping using quadrupeds is a challenging problem, that combines highly dynamic locomotion with precise and fast non-prehensile object (ball) manipulation. The robot needs to react to and intercept a potentially flying ball using dynamic locomotion maneuvers in a very short amount of time, usually less than one second. In this paper, we propose to address this problem using a hierarchical model-free RL framework. Image Credits: Hybrid Robotics Effectively, the robot needs to lock into a projectile and maneuver itself to block the ball in under a second. The robot’s parameters are defined in an emulator, and the Mini Cheetah relies on a trio of moves — sidestep, dive, and jump – to block the ball on its way to the goal by determining its trajectory while in motion. To test the efficacy of the program, the team pitted the system against both a human component and a fellow Mini Cheetah. Notably, the same basic framework used to defend the goal can by applied to offense. The paper’s authors note, “In this work, we focused solely on the goalkeeping task, but the proposed framework can be extended to other scenarios, such as multi-skill soccer ball kicking.”

Stability AI, the startup behind Stable Diffusion, raises $101M • ZebethMedia

By zebethcontrolPosted on October 17, 2022Posted in Robotics & AI, StartupsNo Comments

Stability AI, the company funding the development of open source music- and image-generating systems like Dance Diffusion and Stable Diffusion, today announced that it raised $101 million in a funding round led by Coatue and Lightspeed Venture Partners with participation from O’Shaughnessy Ventures LLC. The tranche values the company at $1 billion post-money, according to a Bloomberg source, and comes as the demand for AI-powered content generation accelerates. Stability AI is the brainchild of Emad Mostaque. Having graduated from Oxford with a Masters in mathematics and computer science, he served as an analyst at various hedge funds before shifting gears to more public-facing works. Mostque co-founded and bootstrapped Stability AI in 2020, motivated both by a personal fascination with AI and what he characterized as a lack of “organization” within the open source AI community. “Nobody has any voting rights except our … employees — no billionaires, big funds, governments or anyone else with control of the company or the communities we support. We’re completely independent,” Mostaque, who serves as Stability AI’s CEO, told ZebethMedia in a previous interview. “We plan to use our compute to accelerate open source, foundational AI.” Stability AI has a cluster of more than 4,000 Nvidia A100 GPUs running in AWS, which it uses to train models including Stable Diffusion. Stability AI plans to make money by training “private” models for customers and acting as a general infrastructure layer. It also offers a platform, DreamStudio, through which its models can be accessed by individual users — Mostaque told Bloomberg that DreamStudio has more than 1.5 million users and Stable Diffusion has more than 10 million daily users “across all channels.” Meanwhile, the open source version of Stable Diffusion has been downloaded more than 200,000 times, according to a press release published by Stability AI this morning. Stability AI claims to have other commercializable projects in the works, including AI models for generating audio and even video. According to Mostaque, the capital from the funding round will support deploying custom versions of Stable Diffusion for users at a larger scale and investing in more supercomputing power. It’ll also be put toward hiring more people, with Mostaque saying he expects to grow to about 300 employees from 100 over the next year. Stability AI has made several high-profile hires recently, bringing on research scientists from Google Brain and futurist and public speaker Daniel Jeffries. Sri Viswanath, a general partner at Coatue, said in a statement: “At Coatue, we believe that open source AI technologies have the power to unlock human creativity and achieve a broader good. Stability AI is a big idea that dreams beyond the immediate applications of AI. We are excited to be part of Stability AI’s journey, and we look forward to seeing what the world creates with Stability AI’s technology.”

Cyberdontics raises $15M for robotic root canals • ZebethMedia

By zebethcontrolPosted on October 17, 2022Posted in cyberdontics, dentist, Hardware, robotics, Robotics & AINo Comments

It’s been more than 20 years since the da Vinci Surgical System received FDA clearance. Pretty incredible when you think about it. Robotic surgery and automation in general have come a long way since then, and a number of companies have entered the lucrative category, focused on all manner of different procedures. Surprisingly, robotic dental procedures have been slow to follow. Let me get this out of the way up front — I’m squeamish about dental procedures. I don’t like thinking about them, don’t like talking about them and certainly don’t like having them. And like many of you reading this, I’m certainly not rushing out to have a robot perform a root canal on me any time soon. I said as much to dentist turned Cyberdontics founder and CEO, Chris Ciriello. Image Credits: Cyberdontics The executive notes that there are two big selling points here from the patient’s standpoint. First is efficacy. He says the system that Cyberdontics is developing will be capable of extremely accurate tooth cutting, down to around 30 microns. The second — and perhaps more important — is speed. “If you’ve had something like a root canal, a crown or any of these types of procedures, where you’re spending an hour or two in the dentist’s chair and you’re spending multiple trips to go back and get it fixed,” he explains, “the idea that you can literally have this robot in your mouth for under one minute and you can be out the door 15 minutes later, is a game changer. For people that really don’t like the dentist, this is a really attractive way to get in and out a lot faster.” The notion was attractive enough to warrant a $15 million Series A for the YC grad. The round, led by dentist chain Pacific Dental Services, will go toward additional R&D and bringing the system to market. The system is supervised by the dentist and, like surgery robots before it, is designed to level access to such procedures amid a dentist shortage. Image Credits: Cyberdontics “Today, a dentist would cut a hole in your tooth and fill the hole with some type of material, whether it’s a crown, a filling, some kind of plastic they squirt in,” says Ciriello. “What we do is scan your tooth, then we virtually create a model of what the tooth will look like after we cut it. Then we can cut your tooth and fabricate a prosthetic at the same time, or we can fabricate the prosthetic in advance of the surgery. Then that piece will fit in just like a puzzle piece, right into the hole we cut.” Cyberdontics “aspirationally” plans to launch its imaging process within the next year, with plans to introduce the robot within the next two, regulator approval depending.

Ambi Robotics secures $32M infusion to deploy its item-sorting robots in warehouses • ZebethMedia

By zebethcontrolPosted on October 17, 2022Posted in ambi, ambi robotics, Bow Capital, Robotics & AI, startup, Startups, tiger globalNo Comments

Ambi Robotics, a startup developing supply chain automation hardware, today announced that it raised $32 million in additional funding led by Tiger Global and Bow Capital, with participation from Ahren and logistics firm Pitney Bowes. Pitney Bowes is a strategic investor in Ambi, having recently inked a $23 million deal with the company to deploy Ambi’s hardware in U.S.-based Pitney Bowes fulfillment centers. The new capital came in the form of a SAFE, or simple agreement for future equity, which grants investors the right to purchase equity in the company at a future date, allowing Ambi to delay negotiations around valuation and terms of investment. CEO Jim Liefer says that it’ll be put toward continuing deployments and installations of Ambi’s tech, expanding the company’s product portfolio and growing engineering, customer support and operations teams headcount. “This additional funding round came together very quickly, spawning from a ‘normal’ company update to our existing investors and partners,” Liefer told ZebethMedia in an email interview. “It sparked interest to further fuel manufacturing and deployments of our current and future categories of AI-powered parcel sorting systems … Just this year, our team has more than doubled and we will continue to add engineering and customer success talent and other areas to keep pace with customer demand for our robotic solutions across their operations.” The co-founders of Ambi — including Ken Goldberg, the chair of the industrial engineering and operations research department at UC Berkeley — years ago discovered clever techniques to train robots in simulation and transfer those learnings to the real world. After a breakthrough on a system called Dex-Net, Goldberg and Jeff Mahler, a former doctoral student, launched the company in 2019, along with other scientists and engineers from UC Berkeley. Dex-Net, short for Dexterity Network, is an AI system that trains on thousands of images of 3D models of objects. Using deep learning, the system scans the data and uses algorithms to learn the best way to pick up the objects. A row of Ambi’s autonomous item-sorting robotic arms deployed to a warehouse floor. Image Credits: Ambi Robotics Ambi’s robotics platform builds on this to automate processes primarily in logistics and fulfillment. The company claims its products, which include robotic arms and the software to run them, can be “taught” to pick and pack millions of unique items while adapting to different packaging (e.g. boxes and envelopes) on the fly. Using “end effectors” like suction cups, Ambi machines pick, scan, insert, place and pack items arranged in mail sacks on fulfillment center floors in tandem with workers. Software running in the background analyzes data on productivity, item dimensions and weights, utilization and more and identifies “pick points” on items in cluttered environments like conveyor belts, totes and bins. Customers pay upfront for Ambi’s robotics units and then pay a monthly subscription cost for use of the software. “The team at Ambi Robotics brings a new way of thinking about traditional problems,” Liefer said. “With advanced tech that can solve a wide range of real-world problems, the team [has] decided to use their expertise to drive the exploding ecommerce industry toward a sustainable supply chain, so the strain of sorting parcels doesn’t rest on the shoulders of our most valuable asset — people.” Liefer says that Ambi’s current focus is deploying the latest generation of its robotics tech, AmbiSort A-Series v3, which features a “soft-touch” end effector that can handle both deformable and rigid items. Ambi claims that warehouse associates can work alongside three to four of these systems to increase the average throughput per employee to over 1,200 items sorted per hour. Ambi competes with Covariant, Nomagic, Soft Robotics, Pickle, Hai Robotics, XYZ Robotics and RightHand, among others, in a favorable investment climate for robotics. According to Crunchbase, more than $17 billion poured into VC-backed robotic startups in 2021 — nearly triple the investment in 2020. In April, Amazon announced that it would create a new $1 billion fund to back companies working in the customer fulfillment, logistics and supply chain sectors. And in May, Walmart expanded its partnership with robotics startup Symbotic to install the latter’s machines into all of Walmart’s distribution centers in the U.S. As of 2019, the global warehouse automation market was worth about $15 billion, according to Statista. That number is expected to double within the next four years, with supply chain executives in an Accenture survey citing automation as one of their top three investment priorities — workers’ concerns about the tech aside. Leifer says that Ambi, for its part, began generating revenue through commercial deployments in October 2020, installing systems prior to the peak holiday buying season. The company is currently in the process of installing 80 parcel-sorting systems while supporting more than 80 “full-stack” sorting systems across 15 sorting hubs. Gregg Zegras, EVP and president of global e-commerce at Pitney Bowes, added in an emailed statement: “Ambi Robotics is an important part of an innovation strategy that is helping Pitney Bowes improve service to our clients and efficiently grow our global ecommerce business. In Ambi Robotics, we see the same commitment to client-led innovation that has helped Pitney Bowes evolve and win in the marketplace for over 100 years. We look forward to continuing to work together to drive innovation in our global ecommerce hubs.” Berkeley-based Ambi, which recently moved into a new HQ, has raised $67 million to date and has more than 50 employees.

Touchlab to begin piloting its robotic skin sensors in a hospital setting • ZebethMedia

By zebethcontrolPosted on October 14, 2022Posted in Hardware, robot, robotics, Robotics & AI, Touchlab, XPRIZENo Comments

Manipulation and sensing have long been considered two key pillars for unlocking robotics’ potential. There’s a fair bit of overlap between the two, of course. As grippers have become a fundamental element of industrial robotics, these systems require the proper mechanisms for interacting with the world around them. Vision has long been a key to all of this, but companies are increasingly looking to tacticity as a method for gathering data. Among other things, it gives the robot a better sense of how much pressure to apply to a given object, be it a piece of produce or a human being. A couple of months back, Edinburgh, Scotland-based startup Touchlab won the pitch-off at our TC Sessions: Robotics event, among some stiff competition. The judges agreed that the company’s approach to the creation of robotic skin is an important one that can help unlock fuller potential for sensing. The XPrize has thus far agreed, as well. The company is currently a finalist for the $10 million XPrize Avatar Competition. The firm is currently working with German robotics firm Schunk, which is providing the gripper for the XPrize finals. Image Credits: Touchlab “Our mission is to make this electronic skin for robots to give machines the power of human touch,” co-founder and CEO Zaki Hussein said, speaking to ZebethMedia from the company’s new office space. “There are a lot of elements going into replicating human touch. We manufacture this sensing technology. It’s thinner than human skin and it can give you the position and pressure wherever you put it on the robot. And it will also give you 3D forces at the point of contact, which allows robots to be able to do dexterous and challenging activities.” To start, the company is looking into teleoperation applications (hence the whole XPrize Avatar thing) — specifically, using the system to remotely operate robots in understaffed hospitals. On one end, a TIAGo++ robot outfitted with its sensors lends human workers a pair of extra hands; on the other, an operator outfitted with a haptic VR bodysuit that translates all of the touch data. Though such technologies currently have their limitations. Image Credits: Touchlab “We have a layer of software that translates the pressure of the skin to the suit. We’re also using haptic gloves,” says Hussein. “Currently, our skin gathers a lot more data than we can currently transmit to the user over haptic interfaces. So there’s a little bit of a bottleneck. We can use the full potential of the best haptic interface of the day, but there is a point where the robot is feeling more than the user is able to.” Additional information gathered by the robot is translated through a variety of different channels, such as visual data via a VR headset. The company is close to beginning real-world pilots with the system. “It will be in February,” says Hussein. “We’ve got a three-month hospital trial with the geriatric patients in the geriatric acute ward. This is a world-first, where this robot will be deployed in that setting.”

Volkswagen to plough €2.4B into vehicle automation in China and form JV with Horizon Robotics • ZebethMedia

By zebethcontrolPosted on October 13, 2022Posted in ADAS, auto, autonomous driving, China, chip, nvidia, Robotics & AI, semiconductor, Transportation, volkswagenNo Comments

Volkswagen is accelerating the pace to automate its electric vehicles for Chinese customers. CARIAD, a wholly-owned automotive software company of the German auto behemoth, intends to set up a joint venture with Horizon Robotics, one of China’s most serious auto chip developers, the company said on Thursday. The German automaker plans to deploy around €2.4 billion to its cooperation with Horizon Robotics, a transaction that’s expected to be completed by 2023 and is subject to regulatory approval. Following the deal, CARIAD will hold a majority stake of 60% in the JV. It wasn’t until 2020 that China moved to ease the rules that had previously barred foreign companies from owning majority stakes in local auto firms. The tie-up comes at a time of global chip shortage and surging semiconductor costs. A handful of automakers are already moving some of their chip production in-house to counter supply chain uncertainties. China’s electric vehicle upstarts Xpeng and Nio have both assembled sizable teams to develop auto-grade chips, according to Chinese tech business publication LatePost. The deal came just weeks after Horizon announced it had received a strategic investment from China’s state-owned automaker Chery Automobile. Together with Horizon Robotics, Volkswagen will be working on full-stack advanced driver assistance systems and autonomous driving solutions for the Chinese market. The goal is to “drive forward the integration of numerous functions on one chip, increasing the stability of the system, saving costs, and reducing energy consumption.” The vision is reminiscent of Nvidia’s recently announced next-generation auto-grade chip that’s designed to unify autonomous driving and in-car technologies. It’s interesting to see Volkswagen forming close ties with a Chinese startup, while Nvidia’s state-of-the-art auto chip is widely recognized as the most cutting-edge in the industry. Given the escalation of U.S. chip limits on China, it won’t be surprising that supply chain diversification is on the mind of VW executives. The question is whether Horizon can deliver something that’s up to par with its American counterpart. In any case, having an on-the-ground partner will likely help VW create more customized solutions for the world’s largest auto market. As Ralf Brandstätter, member of the management board of Volkswagen AG for China, remarks in a statement: “Localized technology development grants the region more autonomy to further expand its position in the dynamic automotive market. Cutting-edge technology comprising the full software and hardware stack, which the new joint venture will develop, will enable us to tailor our products and services even faster and more consistently to the needs of our Chinese customers. Teaming up with Horizon Robotics will allow Volkswagen to accelerate the development of automated driving solutions as part of our NEW AUTO strategy and drive the repositioning of our China business.”

The last mile • ZebethMedia

By zebethcontrolPosted on October 13, 2022Posted in Actuator, Amazon, Foxglove, Robotics & AI, Starship, viam roboticsNo Comments

I don’t love devoting the first several paragraphs of this newsletter to Amazon every week, but no one is making waves — both good and bad — in the robotics space quite like the little mom-and-pop bookseller from Seattle, Washington. This is one of the bad weeks. It’s a story about what happens when your high-profile pilot doesn’t turn out as planned. Failure is always an option. It’s not a good option, and it’s certainly not the option anyone is hoping for, but to suggest it’s not an option is really just a fundamental misunderstanding of what the word “option” means. Life isn’t a motivational poster dressed up as a LinkedIn post — it’s life, and failure is sitting around like a teenager loitering in the 7-Eleven parking lot. It could be a blessing, it could be a curse, but it is never, under any circumstances, not an option. Last week, Amazon confirmed reports that it has scaled back real-world piloting for its last-mile delivery robot, Scout. The ~400-person team will mostly scatter to the wind. A few will remain with the (not entirely dead) project and still others will fill suitable roles inside the company. Amazon tells ZebethMedia: During our limited field test for Scout, we worked to create a unique delivery experience but learned through feedback that there were aspects of the program that weren’t meeting customers’ needs. As a result, we are ending our field tests and reorienting the program. We are working with employees during this transition, matching them to open roles that best fit their experience and skills. Image Credits: Amazon So, what to make of failure in this case? For starters, I’d point to the ups and downs (so to speak) of Amazon Prime Air. The drone project was hit with layoffs during a reorg of the project. However bearish you might (understandably) be about drone deliveries, it’s since made progress, taking baby steps with a smattering of real-world test pilots. Even so, it’s hard not to view the Scout situation as a potential bellwether for delivery robots in general. Amazon is uniquely positioned to make them work, as the world’s largest retailer, which has already found a fair bit of success in the robotics space — primarily through fulfillment automation. It also has more money than god. It would have been easy to continue pumping money into the project. Have you encountered a delivery robot in the wild? — Brian Heater (@bheater) October 12, 2022 There’s a good chance, however, that Scout was simply in the crosshairs of some corporate belt-tightening. Sure, Amazon is fine to toss a few billion here and there for acquisitions like iRobot, but newish CEO Andy Jassy is taking it upon himself to make some cuts to improve Amazon’s bottom line as it faces economic headwinds just like the rest of us. It’s being seen in different spots across the org, and all the robotic vision in the world couldn’t keep Scout from running into this specific obstacle. Starship delivery robots at UCLA campus on January 15, 2021. Image Credits: Starship/Copyright Don Liebig/ASUCLA This space continues to be an interesting one to watch. There’s plenty of VC being pumped into it, and there are a lot of reports around new partnerships. This week Starship announced a partnership with Grubhub that brings its delivery bot to a number of college campuses across the U.S. The list starts with University of Kentucky; the University of Nevada, Las Vegas; Wayne State University; Southern Methodist University; and Fairfield University, with eight or nine more schools being added by end of year. Starship CCO Ryan Tuohy tells ZebethMedia: We have just launched “Delivery by Starship” with Grubhub and we’re in multiple discussions with other partners to offer our world-leading robot delivery experience as a B2B delivery-as-a-service solution. Delivery by Starship integrates into retailers’ existing platforms to make food delivery more sustainable and efficient. Short of a crystal ball, it’s hard to know how all of this will shake out. There are so many moving parts, too many places, too much regulation to consider to accurately predict things five or 10 years down the road. I remain both curious and skeptical about the efficacy around these machines, including how they’ll deal with the ever present threat of things like stairs. Certainly some of these work fine when supervised by a human. And what of teleoperation? It’s become something of a dirty word in a category obsessed with autonomy. The money is certainly there, and vendors are more than happy to partner with these companies. At very least, it’s an indicator to customers and shareholders that you’re looking toward the future. In a world where Amazon has made same and next day delivery the default, more automation could help take some of the onus off humans to kill themselves for quotas. So where is delivery’s Amazon moment? And if Amazon can’t deliver it, who will? Image Credits: Viam Robotics I visited Viam Robotics’ offices last week. Two notes: It’s a big, cool space with a great view of Lincoln Center (this is, admittedly, the less relevant of the two points). The company just rolled out a better beta of its cloud-based robotics tools kit. There are a number of companies pushing to lower the barrier of entry for industrial robotics deployment. It’s exciting to see, though, in our conversation, CEO Eliot Horowitz pushed back on the notion that we’re ready for a low- or no-code solution right now. He told me: Dreamweaver was, in some ways, ahead of its time. If you look at Webflow or Squarespace, they’re kind of doing what Dreamweaver was doing, but Dreamweaver came out at a time when the backends weren’t ready for a product of its nature. It was really just a product ahead of its time. The e-commerce space wasn’t ready for no-code. I think robotics is in the same place. The benefit of a low-code solution, if it worked, would be great. I just think it’s