Factories and warehouses have been two of the largest robotic markets in recent years, as machines, if only to a limited extent, adopt secular processes to speed up work and free people for other, more complex tasks. Now a startup from Poland, which is expanding the possibilities of these robots, is announcing financing, a sign not only for the development of robot technology, but also for the growing demand for more automation, especially in the world of logistics and fulfillment.
Nomagic, which has enabled a robotic arm to identify, pick up, and then pack an item from an unordered selection, today announced that it raised $ 8.6 million, one of the largest seed rounds ever for a Polish startup. Led by Khosla Ventures and Hoxton Ventures, DN Capital, Capnamic Ventures and Manta Ray, all former Nomagic supporters, also participated.
There are a number of robotic arms on the market today that can be programmed to pick up and drop items from point A to point B. However, we are only seeing a new wave of companies that are focused on making these environments a reality. Limitations of these weapons: They can only work if the items have already been "ordered" in a predictable way, e.g. B. on the assembly line. This means that the fulfillment of online orders, for example, is usually carried out by people.
Nomagic has integrated a new level of computer vision, machine learning and other AI-based technologies to improve the capabilities of these robotic arms. Robots powered by their technology can successfully select objects from an "unstructured" group of objects – that is, not a production line, but possibly a different box – before they are picked up and placed elsewhere.
Kacper Nowicki, former Googler CEO of Nomagic, who founded the company together with Marek Cygan (formerly Climate Corporation) and Tristan d & # 39; Orgeval (academic), found that some work had been done on the problem of unstructured objects and objects. Industrial robots – in the US, some live implementations are taking shape, one of which, Covariant, recently ended stealth mode – it was largely a "missing piece" in terms of the innovation that was done to make logistics and processing more efficient ,
That said, there have been few issues with major commercial launches of the technology that opened up an opportunity in a huge market: fulfillment services are expected to be a $ 56 billion market by 2021 (currently the U.S. is the largest) Region, estimated to be between $ 13.5 billion and $ 15.5 billion).
"If every product were a tablet or a phone, you could automate a normal robotic arm for picking and packing," said Nowicki. "But if you have something else, say something in plastic or a really large variety of products, then the problems are there."
Nowicki was a long-time googler who moved back to Poland from Silicon Valley to build the company's first engineering team in the country. During his time at Google, Nowicki worked in the areas of Google Cloud and Search, among others. However, he also followed the AI developments at Google's DeepMind subsidiary and decided to tackle a new problem for his next challenge.
His interest underscores what has been a kind of artificial intelligence fork in recent years. While some of the earliest implementations of the principles of AI actually happened on robots, a lot of robot hardware seems to be clunky and even out of date these days, while the focus of AI has shifted more to software and "non-physical" systems that are aimed at it Replication and improvement of human thinking. Even the word “robot” is used just as frequently in the phrase “robot process automation” today, which actually has nothing to do with physical robots, but with software.
"Many AI applications are not as appealing," Nowicki merely remarked (although Nowicki did not explain it, DeepMind was confronted with a lot of controversy about its own work, particularly in areas such as healthcare). “But improvements to existing robot systems through machine learning and computer vision so that they can be used in unstructured environments caught my attention. There was so little automation in physical systems, and I think it's a place where a lot will change. "
Interestingly, the company concentrates on hardware, but does not actually build hardware itself, but works on software that can be executed on the currently most popular robot arms on the market in order to make it “more intelligent”.
"We believe that most of AI's intellectual property is in the software stack, not the hardware," said Orgeval. "We see it as a mechatronics problem, but even there we believe that this is primarily a software problem."
Having Khosla as a supporter is remarkable, as much of the VC's successful investment has so far been made in North America. Nowicki said he had a connection to the company through his time in the Bay Area, where Vinod Khosla supported a startup before Google (which went bankrupt in one of Dotcom's downturns).
While there is an opportunity for Nomagic to implement his idea globally, Khosla is initially interested in closer collaboration at home, where Nomagic is already working with external logistics and fulfillment providers as well as retailers such as Cdiscount, a French Amazon provider. Soup-to-Nuts online marketplace.
"The Nomagic team has made significant progress since it was founded in 2017," said Sven Strohband, managing director of Khosla Ventures, in a statement. "There is a tremendous opportunity in the European warehouse robotics and automation market, and NoMagic is well positioned to gain some of that market share."
Warsaw, Poland – February 4, 2020 – Nomagic, the provider of intelligent placement robots for warehouses, today announced the completion of a $ 8.6 million seed investment round led by Khosla Ventures. The round is one of the largest starting rounds for a Polish startup. Hoxton Ventures (London) led the round with existing investors DN Capital (London), Capnamic Ventures (Cologne) and Manta Ray (London).
"The Nomagic team has made significant progress since it was founded in 2017," said Sven Strohband, managing director of Khosla Ventures. "There is a tremendous opportunity in the European warehouse robotics and automation market, and NoMagic is well positioned to gain some of that market share."
Based on the premise that warehouse fulfillment requires repetitive manual tasks that are becoming increasingly difficult to find operators, Nomagic develops AI-based robotic arm solutions to reliably pick and place millions of different products. Your intelligent robots can determine how to select never-before-seen products and detect rare anomalies, such as: For example, robots can select two items at the same time. In 2019, Nomagic deployed its solution at Cdiscount, the leading French e-commerce platform, to build the first fully automated packaging line for e-commerce.