Industrial automation is already streamlining the manufacturing process, but first those machines must be painstakingly trained by skilled engineers. Industrial robotics giant Fanuc wants to make robots easier to train, therefore making automation more accessible to a wider range of industries, including pharmaceuticals. The company announced a new artificial intelligence-based tool at TechCrunch’s Robotics + AI Sessions event today that teaches robots how to pick the right objects out of a bin with simple annotations and sensor technology, reducing the training process by hours.
Bin-picking is exactly what it sounds like: a robot arm is trained to pick items out of bins and used for tedious, time-consuming tasks like sorting bulk orders of parts. Images of example parts are taken with a camera for the robot to match with vision sensors. Then the conventional process of training bin-picking robots means teaching it many rules so it knows what parts to pick up.
“Making these rules in the past meant having to through a lot of iterations and trial and error. It took time and was very cumbersome,” said Dr. Kiyonori Inaba, the head of Fanuc Corporation’s Robot Business Division, during a conversation ahead of the event.
These rules include details like how to locate the parts on the top of the pile or which ones are the most visible. Then after that, human operators need to tell it when it makes an error in order to refine its training. In industries that are relatively new to automation, finding enough engineers and skilled human operators to train robots can be challenging.
This is where Fanuc’s new AI-based tool comes in. It simplifies the training process so the human operator just needs to look at a photo of parts jumbled in a bin on a screen and tap a few examples of what needs to be picked up, like showing a small child how to sort toys. This is significantly less training than what typical AI-based vision sensors need and can also be used to train several robots at once.
“It is really difficult for the human operator to show the robot how to move in the same way the operator moves things,” said Inaba. “But by utilizing AI technology, the operator can teach the robot more intuitively than conventional methods.” He adds that the technology is still in its early stages and it remains to be seen if it can be used during in assembly as well.
Continuing its quest to become the Amazon Web Services for biomanufacturing, href=”http://ginkgobioworks.com/”>Ginkgo Bioworks has launched a new spinoff called Motif Ingredients with $90 million in funding to develop proteins that can serve as meat and dairy replacements.
It’s the second spinout for Ginkgo since late 2017 when the company partnered with Bayer to launch Joyn Bio, a startup researching and developing bacteria that could improve crop yields.
Now, with Motif, Ginkgo is tackling the wild world of protein replacements for the food and beverage industry through the spinoff of Motif Ingredients.
It’s a move that’s likely going to send shockwaves through several of the alternative meat and dairy companies that were using Ginkgo as their manufacturing partner in their quest to reduce the demand for animal husbandry — a leading contributor to global warming — through the development of protein replacements.
“To help feed the world and meet consumers’ evolving food preferences, traditional and complementary nutritional sources need to co-exist. As a global dairy nutrition company, we see plant- and fermentation-produced nutrition as complementary to animal protein, and in particular cows’ milk,” said Judith Swales, the Chief Operating Officer, for the Global Consumer and Foodservice Business, of Fonterra, an investor in Ginkgo’s new spinout.
To ensure the success of its new endeavor Ginkgo has raised $90 million in financing from industry insiders like Fonterra and the global food processing and trading firm Louis Dreyfus Co., while also tapping the pool of deep-pocketed investors behind Breakthrough Energy Ventures, the climate focused investment fund financed by a global gaggle of billionaires including Marc Benioff, Jeff Bezos, Michael Bloomberg, Richard Branson, Bill Gates, Reid Hoffman, John Doerr, Vinod Khosla, Jack Ma, Neil Shen, Masayoshi Son, and Meg Whitman.
Leading Ginkgo’s latest spinout is a longtime veteran of the food and beverage industry, Jonathan McIntyre, the former head of research and development at another biotechnology startup focused on agriculture — Indigo Ag.
McIntyre, who left Indigo just two years after being named the company’s head of research and development, previously had stints at Monsanto, Nutrasweet, and PepsiCo (in both its beverage and snack divisions).
“There’s an opportunity to produce proteins,” says McIntyre. “Right now as population grows the protein supply is going to be challenged. Motif gives the ability to create proteins and make products from low cost available genetic material.”
Ginkgo, which will have a minority stake in the new company, will provide engineering and design work to Motif and provide some initial research and development work on roughly six to nine product lines.
That push, with the financing, and Ginkgo’s backing as the manufacturer of new proteins for Motif Ingredients should put the company in a comfortable position to achieve McIntyre’s goals of bringing his company’s first products into the market within the next two years. All Motif has to pay is cost plus slight overhead for the Ginkgo ingredients.
“We started putting Motif together around February or March of 2018,” says Ginkgo co-founder Jason Kelly of the company’s plans. “The germination of the business had its inception earlier though, from interacting with companies in the food and beverage scene. When we talked to these companies the strong sense we got was if there had been a trusted provider of outsourced protein development they would have loved to work with us.”
The demand from consumers for alternative sources of protein and dairy — that have the same flavor profiles as traditional dairy and meats — has reached an inflection point over the past few years. Certainly venture capital interest into the industry has soared along with the appetite from traditional protein purveyors like Danone, Tyson Foods, and others to take a bite out of the market.
Some industry insiders think it was Danone’s 2016 acquisition of WhiteWave in a $12.5 billion deal that was the signal which brought venture investors and food giants alike flocking to startups that were developing meat and dairy substitutes. The success of companies like Beyond Meat and Impossible Foods has only served to prove that a growing market exists for these substitutes.
At the same time, solving the problem of protein for a growing global population is critical if the world is going to reverse course on climate change. Agriculture and animal husbandry are huge contributors to the climate crisis and ones for which no solution has made it to market.
Investors think cultured proteins — fermented in tanks like brewing beer — could be an answer.
Photograph: David Parry/EPA
“Innovative or disruptive solutions are key to responding to changing consumer demand and to addressing the challenge of feeding a growing world population sustainably,” said Kristen Eshak Weldon, Head of Food Innovation & Downstream Strategy at Louis Dreyfus Company (LDC), a leading merchant and processor of agricultural goods. “In this sense, we are excited to partner with Motif, convinced that its next-generation ingredients will play a vital role.”
The investment firm has been busy placing bets across a number of different biologically based solutions to reduce the emissions associated with agriculture and cultivation. Pivot Bio is a startup competing with Ginkgo’s own Joyn Bio to create nitrogen fixing techniques for agriculture. And earlier this month, the firm invested as part of a $33 million round for Sustainable Bioproducts, which is using a proprietary bacteria found in a remote corner of Yellowstone National Park to make its own protein substitute.
For all of these companies, the goal is nothing less than providing a commercially viable technology to combat some of the causes of climate change in a way that’s appealing to the average consumer.
“Sustainability and accessible nutrition are among the biggest challenges facing the food industry today. Consumers are demanding mindful food options, but there’s a reigning myth that healthy and plant-based foods must come at a higher price, or cannot taste or function like the animal-based foods they aim to replicate,” said McIntyre, in a statement. “Biotechnology and fermentation is our answer, and Motif will be key to propelling the next food revolution with affordable, sustainable and accessible ingredients that meet the standards of chefs, food developers, and visionary brands.”
If you’re looking for the next unicorn in Southeast Asia, Zilingo might just be it. The 3.5-year-old e-commerce company announced today that it has raised a Series D round worth $226 million to go after the opportunity to digitize Asia’s fashion supply chain.
This new round takes Zilingo to $308 million from investors since its 2015 launch. The Series D is provided by existing investors Sequoia India, Singapore sovereign fund Temasek, Germany’s Burda and Sofina, a European backer of Flipkart -owned fashion site Myntra. Joining the party for the first time is new investor EDBI, the corporate investment arm of Singapore’s Economic Development Board.
Zilingo isn’t commenting on a valuation for the round, but a source with knowledge of the deal told TechCrunch that it is ‘a rounding error’ away from $1 billion. We had heard in recent months that the startup was getting close to unicorn status, so that is likely to come sooner or later — particularly given that Zilingo has made it to Series D so rapidly.
Raising more than $300 million makes Zilingo one of Southeast Asia’s highest-capitalized startups, but its meteoric growth in the last year has come from expansion from consumer e-commerce into business-to-business services.
CEO Ankiti Bose — formerly with Sequoia India and McKinsey — and CTO Dhruv Kapoor first built a service that capitalized on Southeast Asia’s growing internet connectivity to bring small fashion vendors from the street markets of cities like Bangkok and Jakarta into the e-commerce fold.
Zilingo still operates its consumer-facing online retail store, but its key move has been to go after b2b opportunities in the supply chain. That’s to say that it is building a network of supply chain pieces — manufacturing, logistics, payments, etc — that it can take to retailers or brands. So, in theory, anyone wanting to get into private labels or fashion selling could use Zilingo as an end-to-end solution to make and source their product.
Revenue grew by 4X over the past year, with b2b responsible for 75 percent of that total, Bose told TechCrunch. She declined to provide raw figures but did say net income is in “the hundreds of millions” of U.S dollar. The company — which has over 400 staff — isn’t profitable yet, but CEO Bose said the b2b segment gives it “a clear pathway” to break-even by helping offset expensive e-commerce battles.
Ankiti Bose and Dhruv Kapoor founded Zilingo in 2015.
The supply chain’s ‘outdated tech’
Moving into the supply chain after building distribution makes sense, but Zilingo has long had its eye on services.
That business-focused push started with a suite of basic products to help Zilingo sellers manage their e-commerce business. Those initially included inventory management and sales tracking, but they have since graduated to deeper services like financing, sourcing and procurement, and a ‘style hunter’ for identifying upcoming fashion trends. Zilingo also widened its target from the long tail of small vendors operating in Southeast Asia, to bigger merchants and brands and even to the fashion industry in Europe, North America and beyond that seeks access to Asia’s producers, who are estimated to account for $1.4 trillion of the $3 billion global fashion manufacturing market.
Zilingo’s goal today is to provide any seller with the features, insight and network that brands such as Zara have built for themselves through years of work.
In Southeast Asia, that means helping small merchants, SMEs and larger retailers to source items for sale online through the Zilingo store. But in Europe and the U.S, where it doesn’t operate an outlet, Zilingo goes straight to the sellers themselves. That could mean retailers seeking wholesale opportunities from Asia or online influencers, such as Instagram personalities, keen to use their presence for e-commerce. Beyond just picking out items to sell, Zilingo wants to help them build their own private labels using its supply chain network.
That rest of the world plan has been on the cards since last year when Zilingo closed a $54 million Series C, but now the next stage of the journey is deeper integration with factories.
“If you think about these factories that make the products, the process isn’t optimized over there,” Bose said in an interview. “The guy or girl running factory likely has no technology, they don’t even use Excel. So we’re going to small and medium factories, increasing capacity utilization, helping to manage payroll, getting loans and other fintech services.”
Kapoor, her co-founder, adds that the fashion supply chain is “is marred by outdated tech.”
“It’s imperative for us to build products that introduce machine learning and data science effectively to SMEs while also being easy to use, get adopted and scale quickly. We’re re-wiring the entire supply chain with that lens so that we can add most value,” he added in a statement.
Zilingo encourages retailers and brands to develop their own private labels by tapping into the supply chain network it has built
AWS for the fashion supply chain
Bose said Zilingo’s early efforts have boosted factory efficiency by some 60 percent and made it possible to develop links to retailers while also enabling factories to develop their own private label colletions, rather than simply churning out unbranded or non-descript products.
A large part of that work with factories is consultancy-based, and Zilingo has hired supply chain experts to help provide quality guidance and perspective alongside the software tools it offers, Bose said.
She compares it, in many ways, to how Amazon conceived AWS. After it built tech to fix its own problems internally, it commercialized the services for third parties. So Zilingo started out offering a consumer-facing e-commerce platform but it is making its sourcing networks open to anyone at a cost — almost like supply chain on an API.
That gives its business a two, if not three, sided focus which spans selling to consumers in Southeast Asia through Zilingo.com — which is present in Thailand, Singapore, Malaysia and Indonesia with the Philippines and Australia coming soon — reaching overseas retailers through Zilingo Asia Mall, and developing the b2b play.
In Southeast Asia, its home market, Zilingo doesn’t pressure its merchants to sell on its platform exclusively — “we don’t mind if they go to Instagram, Lazada, Tokopedia and Shopee,” Bose said — but in the U.S. it doesn’t have a go-to consumer outlet. It’s possible that might change with the company considering potential partnerships, although it seems unlikely it will launch its own consumer play.
That’s helped it avoid tricky times for specialist e-commerce services, which battle tough competition, pricing wars and challenging dynamics, and instead become one of Southeast Asia’s highest-capitalized startups. The company’s U.S. plan is ambitious, and it is taking longer than expected to get off the ground, but that makes it a startup that is worth keeping an eye on in 2019. It’s also an example that the startup journey is not defined since, in some cases, the biggest opportunities aren’t presented immediately.
Hard to believe, but we’re only a few months out from the next TC Sessions: Robotics. As we get ready for our third year, take a trip down robotic memory lane with these highlights from last year’s big event.
We’ll be returning to UC Berkeley’s Zellerbach Hall in April, this time with an added focus on artificial intelligence. Last week we announced that computer science professor Hany Farid and VC/Playground global co-founder Peter Barrett will be joining us at the event, and now we’ve got a couple more big names to share with you.
Melonee Wiseis the CEO of collaborative warehouse robotics company Fetch. She previously worked at influential Bay Area robotics startup Willow Garage, where she helped develop the ROS (Robotic Operating System), the PR2 and TurtleBot. Wise has received numerous awards, including MIT Technology Review’s TR35, and was named a 2018 World Economic Forum Technology Pioneer.
Anca Dragan is an assistant professor at UC Berkeley’s EECS (Electric Engineering and Computer Sciences) department, with a focus on human-robotic interaction. Her team explores the fields of autonomous vehicles, manufacturing and assistive robotics. Dragan is a co-founder of the Berkeley AI Research Lab and has received a Sloan Fellowship, MIT TR35, Okawaand NSF CAREER awards.
A former Magic Leap engineer believes the problem with most consumer-facing augmented headsets on the market is their bulky size.
“You wouldn’t want to wear them for more than one hour,” Xu Chi, founder and chief executive officer of Nreal told me as he put on a bright orange headgear that looked just like plastic Ray-Ban shades. Called Light and powered by Qualcomm’s Snapdragon processor, Nreal’s first-generation mixed reality glasses officially launched at Las Vegas’ tech trade show CES this week.
With a light-weight play, the two-year-old Chinese startup managed to bring in some big-name investors. Aside from debuting Light, Nreal also announced this week that it has raised $15 million in total funding to date. The proceeds include a Series A from Shunwei, the venture fund that Xiaomi founder set up, Baidu’s video streaming unit iQiyi, investment firm China Growth Capital, and others. According to Xu, R&D is his company’s biggest expense at this stage.
The financial injection bears strategic significance to Xiaomi and iQIYI. The former is best known for its budget smartphones but its bigger ambition lies in an Apple Home-like ecosystem that surely welcomes portable MR headsets. IQiyi, on the other hand, already has a channel dedicated to virtual reality, which is meant to immerse the end user in a completely digital environment. MR content may just be around the corner to provide an interactive experience of the real world.
Taking money from Shunwei rather than straight from Xiaomi is a thought-through choice. Xiaomi has backed hundreds of manufacturers to gain control over supply chains. Its portfolio companies, in turn, get access to Xiaomi’s retail channels, but they make comprises on various fronts such as product design and pricing.
Xu doesn’t want his freshly minted business to lose independence. “We don’t want to pick sides. We want to be able to work with Oppo and a whole lot of other brands. We want to be compatible with a wide range of devices — smartphones, laptops, PCs, and so on,” said the founder.
Founder and CEO Xu Chi holding Nreal Light’s glasses and chipset. Photo: Nreal
In early 2017, the Chinese entrepreneur started Nreal with his cofounder Xiao Bing, an optical engineer. The brand “Nreal” conveys the partners’ vision to bring users to spaces that fall between the real and unreal. Xu, who spent years working and studying in the US, decided to pursue his ideas back on his homeland for easier access to supply chains.
“We are combining our technological know-how from overseas with great resources in China’s manufacturing industry,” the founder said of his firm’s edge.
The 85-gram (about 3-ounce) Nreal Light isn’t as featherweight as regular glasses but it’s a significant improvement from the biggies it’s going after — Magic Leap One and Microsoft’s HoloLens. Nreal was able to shrink its gadget size because it uses a display solution that requires fewer cameras and sensors than its peers, Xu explained.
Furthermore, Nreal is fixated on the consumer market from the outset, unlike its bigger rivals which, in Xu’s words, are “building gadgets for the next five or even ten years.”
“They want to disrupt everything from cell phones, computers to televisions. They are not necessarily oriented towards consumers,” Xu added.
The smart glasses come in a variety of colors. Photo: Nreal
When it comes to performance, Light claims its display has a 52-degree field of view and a 1080p resolution, which my human eyes weren’t able to verify when I wore it to play an interactive shooting game. That said, I did experience minimum dizziness and latency on Light, as the company promised.
The only irritating part was I started to feel the weight of the specs on my nose bridge a few minutes into my session. Xu assured me that what I tried on was a prototype and that an assortment of nose pads and lenses for different facial features will be available. The glasses also come in a variety of flashy coral colors.
Where does that price tag leave Nreal in terms of profitability? It’s a matter of what kind of consumer hardware Nreal wants to become. “Do we want to be Apple or Xiaomi?” The founder asked himself rhetorically. He’s sure of one thing: As the MR industry matures in China, production costs will also come down. The company is already mulling its own factory so as to beef up supply chains and reduce costs, according to Xu.