First mention must go to Mike Dooley and Labrador Robotics, winner of the inaugural Engelberger Foundation Grant and one of the only commercially viable home robots for the ‘aging in place’ demographic that exists. As you can see from the picture, the Labrador ‘Retriever’ is an affordable autonomous mobile robot capable of lifting and lowering table/shelf height and of retrieving pallets of drinks or medications or other things from countertops and even refrigerators. I’ve been a big fan since I first met the team and they are going from strength to strength.
Second mention to another favorite of mine, Seismic, also a commercially viable ‘aging in place’ solution that is very very relevant for the aging workforce as well. In the not too distant future these ‘wearable robots’ or ‘powered clothing’ will become a workplace essential in order to minimize injury (while also increasing productivity). I imagine that insurance, workers compensation and disability law will eventually make it compulsory to use assistance or augmentation for physical tasks in the workplace.
And the winner of the Cowen Startup Challenge and a $10,000 check at Automate 2022 is…. Covision Quality! Congratulations to all the startups involved. There was quite a difference in startup maturity levels which can make judging an event like this an apples vs oranges problem, but it’s a great opportunity to meet some fresh new robotics technology companies. Here’s an overview of the companies that presented, in order of appearance.
Aivero – Martin Svangtun, CEO
“Cheap robots do not equal Industry 4.0.” Aivero is from Norway and focused on homeshoring, providing robotics to SMEs through the cloud based Aivero platform. The Aivero platform is browser based, accessible by phone or by other vendors, and allows robot controlling for rapid task changeover and cell setup within minutes. Current customers incl NIH, US FDA, Aarbakke, Glamox, Telia and it’s a RAAS pricing plan.
Covision Quality – Franz Tschimben, CEO
North Italy. Automating visual inspection with AI. Metal industry, packaging, printing are primary customers, and existent solutions not appropriate – Cognex and Keyence on average 4 weeks to program and bottleneck finding the experts capable of programming visual inspection. Also visual inspection not always satisfactory. 20% of revenue in manufacturing is ‘cost of quality’ which is a $2.3 trillion industry in the US alone. Up to 30% of all defects are not found, so there is a big gap still in the inspection industry.
Covision Quality provides a real time detect detection production line dashboard with another level of management level decision making. Strength is unsupervised learning, making sense out of hundreds of unlabelled images, which works really well for companies with many many products, or lacking more than a few hundred labelled images. Our software can train itself. and be tested before deployment. Work with quality vision partners – Quality in – Quality out.
Software deployed in 1 hour, accuracy 4x greater than competition. Good customer feedback. from Italy, Germany and US. Scaled over the last year and half in spite of challenging times. Free consultation, $15k for first installation two months but much faster in future, then SaaS per year per object pricing. Want to be the SaaS company for visual inspection of everything manufactured.
FLX Solutions – Matthew Bilsky, CEO
Inventor, contractor, mech eng professor. Logistics has last mile, inspection and maintenance has the last meter problem, where experts and engineers can not access the spaces they need to go.
Each link is identical and packed full of sensors, combining multiple links, patent pending. Multiple end effectors can be added from painting, spraying, vacuuming, drilling etc. Example use case – inspecting and planning ceiling BIM including pipes present or missing. This is valuable for energy efficiency updates. Can then remove debris or add insulation.
Can be used on the ground, or without need to enter confined spaces. Greater health and safety for workers. Highly adaptable. Multi trillion markets, construction and BIM market, can extend to agriculture, shipping etc. Partnered with Gentler, Lamarr.ai, Skyline capital builders, one of three DOE prize winners. Product roadmap towards seed round at end of 2022 to increase seed units out with pilot customers and drive price point down to open up to economies of scale in production and turnkey solutions. Self driving and data monetization are future land marks for FLX.
INNOCISE – Henrik Ollmann, Head of Technology
We turn robots into geckos. 3yr old company, spin out. Innovative and precise handling solutions inspired by the gecko. Tiny micro structures allow the amazing handling a abilities without power supply. Strong attachment without need for great power. Great potential for automation. Our grippers have many very tiny hairs, reducing Co2 footprint reduction by energy-free handle without compressed air infrastructure of pneumatic or vacuum gripping. This opens up challenging environments such as clean rooms, vacuum production. and manipulation of fragile objects or surfaces from um2
2021 German Innovation Award for ADHESO, also 50% female employees. 2022 German Innovation Award for Climate and Environment. Over 100 international customers. Pharmaceutical, battery and fuel cell assembly, packaging and clean rooms, glass fiber handling and more. Savings of uptown 2 mio. Euro through no compressed air infrastructure and reduction of CO2 footprint. Unique selling point, our grippers are made of very soft polymers, adaptive, sensitive, tailored to your exact needs, sustainable in requiring no additional energy. Use your existing machinery, just eliminate your vacuum grippers.
InOrbit – Florian Pestoni, CEO
Leader in Robot Operations of robots at scale – RobOps@Scale . Global supply chains are broken. Labor challenges (Korn Ferry world is faction 8.5$T by 2035). Delivery expectations increased (shot of Tiny Mile) now our delivery companies are promising 10 minute delivery. Smart robots can help with all these problems, but software infrastructure is lacking.
If you come from the software or cloud world, then DevOps is what makes it all work. We are bringing those concepts to the world of robotics. We are thinking of fleets, not individual robots. Not every end user is at the same point in their maturity of robots. Single robot, single site through to multi robots multi sites. We want to maximize the potential of every robot. Our solution is a distributed architecture that can handle running on the robot, on the edge to cloud and control. API and Agent.
Everything from knowing what’s going wrong with your robot, to orchestrating deployments and robots across different sites and different vendors. The four O’s of InOrbit. Funding/venture baked to become a catalyst for all robots. Key customers ranging from startups to giants like Karcher and Omron.
Neurocle Inc – Hongsuk Lee, CEO
Auto Deep Learning Software. Making Deep Learning more Accessible – Awards from Vision Systems conf 2021 and 2022 also Gartner Cool Vendor. Very brief founder slide – missed it totally – very different approach from Asian company! Too many slides with text going way too fast through them. Not calibrated to audience. Auto DL Algorithm automatically optimizes and adjusts the various parameters for DL model. ie. 1 month using Neurocle compared to 1 yr DIY or current process. What is case study?
GUI based no code software. Workflow video. Easy training data interface. checking result of . Three types of NEURO- software depending on need/skill. Founded 2019. Korean customers, now sales pool in 10 countries , incl panasonic, etc. Can play the role of skilled deep learning engineer and tune for fast deployment on your own site.
Ninox 360 – Peter Abeles, CEO
Reducing cost of manual processes, without time and money costs of implementing automation. Poka-Yoke approach, continuous feedback to technicians enabling them to find mistakes at the source, and produce automatic metrics and documentation, more effective than traditional training. Using vision systems to introduce. Example being headlight calibration. Customer complaints, engineer confusion over ‘correct’, open loop and evaluated on speed alone. Solution. Camera system to verify correctness, car could only progress if calibrated. Much cheaper than automated solution. A way of providing feedback to humans?
Other case studies – camera calibration and drone testbed. NINOX 360 specializes stability estimation so you know when to trust measurements. Ie humans what to line up directly above, whereas vision system prefers offset in order to avoid singularity.
Selling bespoke vision solutions, which can include Poka Yoke. Any market where industry used humans to perform a task that uses vision. Customers include Tesla Accel Robotics, Hebi Robotics. Smart sensors as a service, integration, customization.
Quotebeam – Nikki Gonzalez, Head of Partnerships
Revolutionize the way that customers find and buy everything on the floor at Automate. The biggest innovation in procurement has been ditching the fax machine in exchange for email in the 1990s. Why can’t we bring consumer level recommendation engines etc for industry. ??
Information is highly silo’d and procurement can take weeks, even just getting a Bill of Materials. Engineering shouldn’t be about scraping the internet for information – often inaccurate, out of date or out of stock.
Quotebeam allows engineers to manage this process through one platform connecting thousands of independent suppliers on a single platform. Distributor network, quoting, ? , ? ,
Build your BOM, with access to tech specs, models etc. See who has stock, availability and supply time. As you add parts to your BOM, QB auto requests quotes and puts items on hold. When you place the order we connect to their ERP systems and no additional data entry is required. Or you can send them a request to join QB and still contact them manually. It’s an Opt In. Tesl, pesto, Coca Col, Foster Farms, Frito Lay. We want to work with everyone from machine shops to multinationals. Why are we the perfect team to reinvent procurement? Our CTO has been working for world’s biggest brands for years. We know what’s needed, how to build it. Everyone on this show floor has expressed need for our service. In last two months/years since we launched we’ve grown 10x. 1000s of customers worldwide
Scalable Robotics – Tom Fuhlbrigge, CEO
Code free CAD free point of ??
13 yrs direct of ABBs corp Research, background MIT etc. Many US patents.
Industrial robotics is too complex. Large industrial application areas like welding, painting and gluing require Cumbersome teach pendants, complex programming languages, or CAD tools. Often all three. Manufacturing is facing a skilled labor shortage. We enable a non skilled worker to train robots, with Scalable’s Teach platform utilizing vision, hueristics, ML. When we have the model of the part we model that in simulation and make sure the robot can complete the task. Then we can autotune the robot program. WE capture the local worker knowledge and break down the last barriers to adopting automation and training workforce, with a quality process that requires no additional programming knowledge which enables you to keep your lower skilled worker on staff. You also save many $k not upgrading your processes for inspection upstream by ensuring quality initial.
Beach head market is welding – huge labor shortage. $7b market global 30% in US and very easy ROI with robots, where they are 3x more efficient. Then do follow on applications.
Simple hardware modules Converting never used a robot to ROI on first day, buying more. Huge market, major OEM partnership, lots of early customers, let’s grow!
VoAIge – ? , CEO
Operating AI for robotics arms. Vision based operating AI for robotic arms. Cool 1938 to 2022 video – basically cobots are just taking us back to 1938 🙂 Fixed infrastructure for industrial robots doesn’t work for logistics industry or many SMEs etc. Vision based reprogrammable operating AI software requires no tech background and works with just about any robot arm. It’s a computer including the vision sensors with preinstalled AI software/US can power any work cell. Our first series is called Deep Space. Can pick unseen complex parts/skus and scale to thousands seamlessly. not fixtures needed… retry unsuccessful grips. Using EOAT cameras allowing you to move around anywhere.
Can handle clutter or complexity in workflow. What do you want to do and where do you want it to go? Basic two fingered grasp shown. Come see the demo in the booth. Capable of dealing with dynamically changing conditions – therefore machine tending is a viable task, ditto complex picking. How do we fit into the market? We add value to everyone. Systems integrators – increase TAM. OEMs – increase usability of low cost arms. End users – obviously AMRs – ….
Working palletizing, bin picking machine tending as three initial use cases. Huge team p/t time interns and 3 key people.
Conclusion
According to the organizers, there were many applications for the competition, so each company who presented is a winner already. It’s especially pleasing to see so many Silicon Valley robotics companies representing! (InOrbit, FLX Solutions, Nineox, Quotebeam in the competition, as well as Seismic and Labrador Robotics)