Growth, Innovation + Investing
from A Round to Apple Inc.

We’ve been delivering authoritative market analysis and perspectives, week-in and week-out for more than a decade. Sign up so we can update you with new releases of A2Apple and our other daily and weekly publications — sent straight to your email.

And make sure to keep a close eye on our research and commentary arm, GSV Tomorrow, where you’ll find original GSV insights, plus curated content revolving around our key Growth Themes. Just complete this form and they’re yours to use and share.

GSV's weekly perspective on Growth, Innovation + Investing...
from A Round to Apple Inc.
Cloud & Big Data
August 5, 2018

MyBOT, YourBOT, Everything’s a Robot

Save to PDF

Market Snapshot

Indices Week YTD

On April 3, 2012, the Seattle Times ran a story titled, “Amazon Warehouse Jobs Push Workers to Physical Limit.” The central character was 51-year-old Connie Milby, who covered more than 10 miles per day in her tennis shoes moving packages from shelf to shelf at an Amazon warehouse in Kentucky.

“As long as my body holds up, I will keep working,” Milby reflected on her $12.50-per-hour job. “But the way it feels, I don’t know how long that will be.” Not exactly the stuff of Fortune’s Best Companies to Work For list.

Today, Amazon “employs” over 100,000 Kiva robots, which now account for nearly 20% of the company’s total workforce. You won’t be reading interviews with them any time soon, but rest assured these sturdy warehouse workers can move 10 miles without breaking a sweat.

What happened? It all started a month before Milby’s interview.

In March of 2012, Amazon announced the acquisition of “material handling” robotics maker Kiva Systems for $775 million. The second largest M&A deal in Amazon’s history, it seemed like a classic Jeff Bezos moonshot.

Kiva made clunky warehouse robots that looked like oversized orange hockey pucks. Wall Street analysts noted that the price tag was a 300% premium to the company’s last private valuation. It was led by a former employee of Webvan, the high-flying delivery service that went belly up in the dot-com bust.

It turns out the Bezos was on to something. Amazon knew what Kiva’s robots were capable of when it acquired the company based on pilots with and By the time Amazon fully implemented the new technology in 2014, the company demonstrated that it could cut the “click to ship” cycle down from the approximately 75 minutes humans required to just 15 minutes. Kiva-enabled warehouses operated with 20% reduced operating costs.

Total Number of Robots Amazon has Deployed in Warehouses

Source: Quartz, New York Times, Bloomberg

Is the robot revolution upon us? You’ll have to ask Alexa. We believe that rapidly changing technology fundamentals, coupled with broader robotics applications across a variety of industries, signal an inflection point for an industry that has not yet seen the Moore’s Law-type scaling that the world has come to expect from other innovation areas.


Science Fiction allows you to picture a future that might otherwise be unimaginable. Not coincidently, many of the great innovators and inventors were science fiction devotees, and you can see in modern devices.   

The iPad and iPhone? Go back to the Starship Enterprise and check out the devices Scotty and Spock were using. The Apple Watch? Dick Tracy had one in 1946.

We’ve seen robots on the silver screen for nearly a century, often in galaxies far, far away. Most have human-like intelligence — just a lot more of it. Some are friendly. Others… well… watch Terminator or the Matrix.

Are these visions crazy? Maybe. But we’re closer to the robots of our imagination than you think. Rapidly improving technology fundamentals, transformational developments in Artificial Intelligence, and accelerating investment activity means that science fiction is quickly becoming science reality.


Source: GSV Asset Management

1. New Technology Fundamentals

Until recently, robotics innovation and production of any significance has been largely limited to a small group of industrial conglomerates in the United States, Japan, and Germany. Six of the ten largest robotics manufacturers are over 100-years old.

Six of the Ten Largest Global Robotics Companies Are Over 100-Years-Old

Source: Yahoo Finance, GSV Asset Management
*Lockheed Aircraft Co, and Glen L. Martin Co. each founded in 1912, merged in 1995; ASEA (est.1883) and Brown, Boveri & Ice (est. 1891) merged in 1988 to form ABB; Northrop Corp. (est. 1939) purchased Grumman Corp. (est. 1930) in 1994; Rockwell International acquired Allen Bradley Co. (est. 1903) in 1985 to form the predecessor to Rockwell Automation
*Market Cap as of August 1, 2018

A key megatrend that is opening the door for new market entrants is the rapid decline of technology costs, from hardware components to computing horsepower.

From 2000 to 2015, computing power increased over 100x, while the cost of computing and storage fell more than 99%. During the same period, the price of bandwidth decreased over 98%. Welcome to a world ruled by Moore’s Law.

Costs of Computing, Bandwidth, and Data Storage, 2000 vs. Today

Source: Deloitte, GSV Asset Management

Smartphones — which didn’t exist until 2006 — are the manifestation of these new technology fundamentals. Today there are nearly three billion smartphone users with computers in their pockets more powerful than the 1969 NASA command center machines that landed men on the moon.

Moore’s Law has driven down the cost of robotic components and enabled entrepreneurs to create machines that are faster, smaller, cheaper, and more powerful. Anki, for example, is a company founded by Carnegie Mellon engineers that is developing toy robots using the same components that power autonomous vehicles. Their flagship product Cozmo — a friendly looking bulldozer that is for all intents and purposes self-aware — would have been unimaginable a decade ago. Anki has raised over $157 million from a syndicate of leading venture capital firms, including Andreessen Horowitz.

Despite the transformative impact of Moore’s Law, technology challenges to scale still remain for entrepreneurial robotics companies.

For large-scale industrial robots, installation can be as expensive as the robots themselves. Autonomous robots remain capacity strained by battery technology that has only improved incrementally since the 1970s. And we haven’t even considered the social stigma associated with machines that are presumed to eliminate human jobs.

Add it all up and the robotics industry appears to be in the age of the Blackberry before Steve Jobs unveiled the iPhone in 2007. Most of the technology components are in place. Robots are heavily used in narrow segments, primarily in the manufacturing sector. With the right design and positioning, the iPhone scale could be around the corner.

2. The Age of “Smart Robotics”

In the old world, robots were deployed to complete precise, repetitive work, in carefully designed settings like an assembly line. Adaptability was neither a capability nor a priority. That’s all about to change.

Deep Learning is a field of Artificial Intelligence where computers “teach” themselves concepts and tasks by crunching large sets of data. It’s a way of getting computers to know things when they see them by producing rules that programmers cannot specify. In other words, it’s a way to develop smart machines at scale much faster than previously imagined.

For example, Facebook’s facial recognition algorithm, Deep Face — which can recognize human faces with 97% accuracy — was created by feeding computers with millions of images of faces. The same principle can be applied to teach a car what a “Stop” sign is, or how to merge onto a highway.


Source: Alphabet, Los Angeles Times

The self-driving car is a prime example of “smart robotics.” Autonomous vehicles couple location information from fixed sensors with maps of the world to navigate. Programs processing optical data then generate instructions for the car’s motor and steering systems. Most importantly, applying advanced Artificial Intelligence approaches like Deep Learning enables self-driving cars to “learn” from their surroundings.

“Cloud Robotics” is another trend accelerating innovation in the robotics sector. As Keith Nosbusch, the former CEO of Milwaukee-based Rockwell Automation observed in a 2014 interview with the Financial Times, “Manufacturing has always been producing a lot of data — but generally it hasn’t been used. But the data can be incredibly useful. When something goes wrong — a pump fails or a valve breaks down — it can be turned into something that tells you why.”


Source: GSV Asset Management

Cloud-based data networks enable engineers to more efficiently design and optimize robotic performance (or, in the case of Deep Learning, robots optimize their own performance). Open source platforms like ROS — Robotic Operating System — have been a boon to researchers and entrepreneurs alike. ROS provides a standard platform for programming robotic components. It also provides libraries and algorithms for vision, navigation, manipulation, and other capabilities.

3. VC Investment Activity

A key indicator of accelerating innovation in the robotics sector is growing interest from venture capital firms. According to CB Insights, since 2010, VCs have deployed approximately $7.8 billion into robotics startups across nearly 1,200 deals. Over $6 billion of these investments have occurred over the last five years.


Source: CB Insights

In the first half of 2018 alone, over $2 billion of funding has been deployed across 132 robotics deals — equal to all investment activity in 2017. Notably, UBTECH, an Artificial Intelligence-powered humanoid robot company producing everything from toys to robo butlers, raised $820 million in a May Series C financing. The round was led by Chinese social media behemoth Tencent. After the investment was announced, UBTECH CEO James Zhou indicated the company would use its war chest to accelerate its vision of bringing robots into the home.


Source: Crunchbase, CB Insights, GSV Asset Management


Corporate investment in robotics startups has accelerated alongside venture capital activity, surpassing $2 billion across more than 100 deals since 2017.


Source: CB Insights

Over the past few years, leading technology companies, including Amazon, Alphabet (Google), Softbank, Foxconn, Qualcomm and GE, have aggressively moved into the field of robotics.


Beyond Amazon’s aggressive integration of Kiva robots and mechanical arms into its warehouses, the company is incubating a variety of robotics initiatives. It is reportedly working on a home robot dubbed “Vesta”, the Roman goddess of home and family. This initiative is housed under Amazon’s Lab126 hardware R&D division — the same group that developed Amazon’s Echo and Kindle devices.


Source: Company Disclosures, GSV Asset Management

It’s unclear what tasks an Amazon robot might perform. According to Bloomberg, “people familiar with the project speculate that the Vesta robot could be a mobile Alexa, accompanying customers into parts of their home where they don’t have Echo devices. Prototypes of the robots have advanced cameras and computer vision software and can navigate through homes like a self-driving car.”

Alphabet (Google)

In just over a year, Alphabet kickstarted the creation of a robotics innovation arm, acquiring eight promising robotics startups beginning in 2013.

The company kept its plans close to the vest, but it appeared intent on applying the technology a broad array of autonomous systems — from self-driving cars to warehouse work, package delivery, and even elderly care.

Key Alphabet Acquisitions in Robotics

Source: GSV Asset Management

But in 2014, Andy Rubin, the founder of Google’s robotics division and the brainchild behind the acquisition spree, left the company to start his own incubator for hardware startups.

In retrospect, it signaled a shift in Alphabet’s strategic priorities.

In 2017, Alphabet sold off two of its robotics acquisitions — Boston Dynamics and Shaft — to SoftBank. The remaining acquisitions have seemingly been left on the back burner, with no commercial application beyond Alphabet’s walls.

Instead, Alphabet appears to be focusing its attention on creating a software layer for robotics, much as Google maps is an enabling layer for autonomous vehicles. The company’s “Robotics at X” initiative, for example, is investigating how machine learning powered by massive data sets can be used to teach robots new skills that will enable them to reliably perform useful tasks in unstructured environments.


Source: Company Disclosures, GSV Asset Management

On the venture side, Alphabet continues to be active in investing in robotics startups through its venture arm, Google Ventures. To date, it has backed companies including Wonder Workshop (educational robots), Abundant Robotics (agriculture robots), Veo Robotics (industrial robots), and Savioke (delivery robots). 


Industrial & Manufacturing

Manufacturers like GM have integrated robots into assembly lines since the 1960s when the automaker installed a hulking 4,000-pound arm called the “Unimate” to attach die castings to car doors.

Today, a new generation of startups is applying advanced robotics to automate a broad range of nuanced physical processes. The most integrated use of robotics can be found in the industrial and manufacturing sectors due to the sheer number of repetitive tasks that can be made more efficient through automation. Bots are increasingly popping up in warehouses and retail settings as well.


Source: IFR, GSV Asset Management

While Amazon’s Kiva robots buzz around busy warehouses, Lowe’s and Walmart are bringing inventory management robots developed by Bossa Nova Robotics to their store aisles. Bossa bots roam the aisles to scan the shelves and catalog products, enabling Walmart employees to focus on providing better service to customers. Once a bot notices that a shelf is low on a product, the machine pings an alert to an employee to restock, optimizing everyone’s time.


Source: Crunchbase, CB Insights, GSV Asset Management

In four years, Taiwanese electronics manufacturer Foxconn has eliminated over 400,000 jobs in favor for manufacturing and warehouse robots. The company already produces 10,000 “Foxbots” annually and intends to aggressively automate its sprawling workforce moving forward. It is estimated that Foxconn employs over one million workers in China alone. The company plans to reach a 30% level of automation by 2020.

In this context, “China” itself might be the key megatrends in industrial and manufacturing robotics. To date, the world’s largest producer of everything from clothes to electronics, has depended on low-cost, low-skill labor. But as economic growth slows and wages continue to rise, manufacturers are aggressively adopting robotic solutions.

Annual Acquisition of Industrial Robots By Country

Source: IFR, National Robot Associations, GSV Asset Management

In 2015, China announced their government-backed “Made in China 2025” plan, which emphasized the nation’s focus on innovating in cutting edge technology industries, including robotics. As a part of this plan, China declared they want to build at least 100,000 industrial robots per year by 2020. The country looks to be on track, deploying over 138,000 robots in 2017 — a more than 50% increase from 2016. Compare that to the United States, whose robot workforce only grew by 5% in the same period.

Transportation + Delivery

A key industry that will be transformed by driving and flying robots is delivery and logistics.

We’ve heard this promise before. In the dot-com bust, on-demand delivery services like Kozmo, Urbanfetch, and Webvan became a symbol of the excess that fueled the crash. The idea then was that the Internet would create efficiencies that offset logistical costs. At its peak (or valley, depending on your perspective), Kozmo was charging $2 per delivery while hemorrhaging money on courier and warehouse expenses (in 1999, Kozmo reported $3.5 million of revenue on $29 million of expenses).

Beginning in 2013, a new crop of on-demand delivery businesses emerged with an eerily similar value proposition, from DoorDash ($721M raised), Instacart ($1B raised) to Postmates ($278M raised), Deliveroo ($860M raised). China’s ($2.3B raised). As private financings surged, Benchmark-backed GrubHub recorded a successful IPO in 2014 and is valued at nearly $12 billion today. China’s was acquired by Alibaba earlier this year for $9.5 billion.

From the Internet, to Smartphones & Marketplaces, to Robots

Source: GSV Asset Management

The key difference is that in 2000, there were only 370 million people on the Internet (roughly 5% of the world’s population), smartphones were a fantasy, and applications off of a platform had not been invented.

Robo deliveries, which eliminate human labor costs, are the next phase. Starship Technologies, for example, conducts door-to-door deliveries with six-wheeled container robots that now operate globally. Starship started by piloting autonomous delivery programs with DoorDash and Postmates. Today, the company runs pilots on college and corporate campuses and aims to deploy 1,000 delivery bots by the end of this year.

Starship’s Door-to-Door Food Delivery Robot

Source: Business Insider

Since it merged with Segway in 2015, Beijing-based Ninebot has been developing and launching transportation robots designed to enable last-mile deliveries. The company raised a $100 million Series C financing in October 2017.


Source: Crunchbase, CB Insights, GSV Asset Management

In 2016, San Francisco-based Zipline began using drones to deliver blood from Rwandan blood banks to rural areas for emergency transfusions. Right now, the country struggles to get blood to remote clinics and can’t predict in advance which blood types they’ll need. Furthermore, 75% Rwanda’s roads are unpaved and often unusable during the rain season, making them impassable to the vehicles that are deployed to make emergency medical deliveries.

Working with UPS and vaccine distributor Gavi, Zipline is serving 21 medical stations throughout Rwanda delivering blood supplies in less than 30 minutes.

Services + “Social”

The first generation of service robots were designed to assist people with repetitive tasks like housecleaning. In 2002, iRobot created the most popular service robot to date, the Roomba vacuum cleaner, which has sold over 15 million units since its launch.

Today, Knightscope robots patrol malls and public areas, coordinating security services and other first responders. Sally the Salad Robot created by Chowbotics, effortlessly creates multi-component salads.


Source: Crunchbase, CB Insights, GSV Asset Management

Anki builds entertainment robots that combine the aesthetic of Pixar’s WALL-E with the functionality of Disney’s Baymax. Its Cozmo product, for example, is a toy robot with built-in facial recognition that is able to develop skills over time through play and use.

But what’s special about Cozmo is its emotional adaptability. It responds to human interactions in very human ways, throwing tantrums and expressing excitement in eerily appropriate settings. Anki recently released Cozmo’s SDK, allowing engineers to build programs off its platform, effectively doing for robotics what iOS and Android did for mobile technology.

Over the past two years, we’ve seen multiple megarounds of financing completed by companies developing social robots, including UBTECH and Rokid, which raised $840 million and $100 million in 2018 respectively.


Source: Crunchbase, CB Insights, GSV Asset Management

A special thanks to Erin Wong, our intern from Carnegie Mellon, for leading the research and development of this week’s piece.


by Luben Pampoulov

Fast Movers

In a few months when we write our “2018 year in review” analysis, there will be one specific area that will stand out — Electric scooters took the world by storm this year! Specifically Lime and Bird made the biggest societal impact, with their Green-White and Black scooters, respectively, becoming the most popular (and coolest) way to move around town.

The fascinating part is that the scooter boom is just a derivative of multiple iterations in urban transport innovation; Uber pioneered the on-demand black car service and the in-app payment, Lyft followed and pioneered peer-to-peer ride sharing, Ofo and Mobike later launched the dock-less bike service, and Bird and Lime applied all of the above to scooters. (Disclosure: GSV owns shares in Lyft).

Lime, which is now 14 months old, recently announced it passed 6 million rides, up from 3 million just two months ago. To put this in perspective, Lyft reached its 6 millionth ride after more than 3 years of operations. It took Uber even longer.

With the meteoric rise of electric scooter (and bikes), many people think the once hyped ride sharing industry is in danger… But the reality is that ride sharing companies continue to grow at impressive rates. Lyft is still growing in the triple digits, Grab continues to expand its leadership position in SE Asia, and Indonesia’s Gojek is following closely on its heels.

Not only is ride sharing penetration in the U.S. still below 30%, and there is room for growth, but the frequency of usage continues to rise too.

Ride Sharing Rides Per Passenger Per Quarter in the U.S. 
(as of Period Ending)

Source: GSV

Last week, Singapore-based Grab raised $2 billion at an $11 billion valuation. The company’s value has increased over 10-fold in just four years time, and the company is expected to generate more than $1 billion in annual net revenue this year.

Similarly, Gojek has been moving fast in grabbing marketshare in its home market Indonesia, and recently also expended into Grab’s territory (after Uber’s defeat).

Interestingly, Grab and Gojek have became what Uber always wanted to be — an all inclusive platform for all types of on-demand services. If you live in Indonesia, you might just have the Gojek app on your phone, and you will be fine… It has the most popular payment service, the biggest food ordering service, groceries, movie tickets, massages, cleaners, cosmetics, auto parts, gifts, etc. Oh yes, you can catch a ride too.

Grab, while larger than Gojek on a ride volume basis, has yet to expand its platform to as many services. 

Pioneer Notes

by Li Jiang

Blockchain for Everyone

A Harmonious World

“I came to the conclusion that we should aspire to increase the scope and scale of human consciousness in order to better understand what questions to ask. Really, the only thing that makes sense is to strive for greater collective enlightenment.” 

— Elon Musk

Since Satoshi’s Bitcoin white paper fired the “starting gun”, innovation in blockchain technology has opened a pandora of opportunities. From Bitcoin, a store of value / currency to Ethereum, enabling smart contracts, we are in the top of the first inning.

To me, the ultimate reward that the blockchain community is working towards is not crypto-kitties or an ETF that passes muster for the SEC, but a way in which we unleash the powerful creative energy of all of humanity.

This requires blockchain because a distributed, trust-less system creates incentives for productive creators while supplying perfect personalized services for consumers.

We not only need greater collective enlightenment, but also greater collective empowerment.

The centralized digital economy has made many people in my neighborhood rich and powerful, but it has done little to move the needle for the rest of the 7 billion people.

In an Open Internet Economy, enabled by very fast blockchain with low latency and negligible fees, creators can directly provide consumers with ideas, content, art, entertainment, financial services, health and wellness guidance, file storage, commerce, etc. For creators, they won’t have to worry about someone taking away their API access or increasing the price of doing business arbitrarily. Furthermore, they don’t have to worry about whether their zip code is 94110 (SF) to find funding. And for consumers, they no longer have to be forced to choose between mediocre services and solutions that aren’t personalized to their own liking.

I think the excitement of the potential of blockchain is not just another way to build an enterprise application or to track diamond supply, which are nice use cases, but to open the floodgate of opportunity to over 7 billion people in the world. The dream of blockchain innovators is nothing short of a new Open Internet Economy where both consumers are able to find products and services that delight them and entrepreneurs can unleash their full potential without gatekeepers and hostile arbiters.

That is a tremendous future that blockchain innovators can make into a reality.

“We cannot solve our problems with the same thinking we used when we created them.”

 — Albert Einstein

Market Update

Week ending August 5, 2018

World Indices

U.S. Indices Snapshot

Valuation P/E Est. P/E/G Price/Sales