Computer Vision: Broadening Vistas of Opportunity for Venture LPs
First Glimpses of the Future Economy
When asked to picture the future, many people will conjure up visions of robots and autonomous machines operated by artificial intelligence. Few stop to imagine that the concept of vision itself might be the source of consequential disruption to the ways that day-to-day commerce is carried out. But according to Michael Suswal, co-founder and COO of Standard Cognition, a TI Platform Management company, the future of commerce will be shaped less by these visions of computers than by computer vision, a term that refers to deep-tech-enabled systems that incorporate visual detection technologies such as cameras and lidar sensors.
On November 12, Suswal joined TI Platform Management for a webinar discussion of Standard Cognition’s innovative computer vision technology and its potential to reshape the global economy. Suswal wasn’t bashful with his claims: “We believe computer vision is going to be the technology that will have a bigger impact than the internet had on the world,” he told the webinar audience: “bigger than the impact mobile had on the world: something on the scale of personal computing.”
Investors share his interest: as of its series B in mid-2019, Standard Cognition had raised $86 million to scale the deployment of its autonomous checkout technology and software to enterprise retailers. The company has grown from seven to 135 employees and has offices in San Francisco, Milan, and Tokyo.
We believe Standard Cognition is poised to become a breakaway success. But it’s also just one among a larger herd of well-funded startups that promise to remake the way business is done around the world—and the way people perform even the most commonplace tasks. Due to the low cost of the hardware required to implement computer vision platforms and the manifold applications for the technology, computer vision–enabled deeptech services like those offered by Standard Cognition are leading the next wave of disruption across entire sectors.
A Case Study
Michael Suswall’s bullish outlook for Standard Cognition is based on the company’s rapid success in deploying its computer vision-based autonomous checkout technology in its own prototype stores. In the Standard convenience store on Market Street in San Francisco, customers enjoy cashier-less checkout—meaning they never have to stand in line—simply by installing the Standard Cognition app. The technology works through a system of cameras located on the store ceiling, which uses computer vision to track the movement of every item in sight. Customers get what they need and leave, receiving a digital receipt for their purchases within 30 minutes of their departure from the store.
The implications of the Standard Cognition camera and SaaS platform are historic. We found that ever since NCR came up with the cash register over 150 years ago, there hasn’t been a breakthrough in retail. Although the rise of online merchandising, the emergence of drop-ship manufacturing, and the collapse of Main Street retail are often mentioned as evidence of breakthrough-grade shifts in retail, brick and mortar retail is still here to stay. The more important shift—and a result of the ease of online shopping—is from individually owned stores and local chains to national and multinational enterprise retailers. This is where software-enabled innovation stands to make enormous strides in reshaping the entire market sector.
Standard Cognition is one of several companies leading the way to this transformation—but unlike its competitors, it doesn’t think of itself as a retailer or a retail vendor. As Suswal explained, retail is only one application of the computer vision technology that Standard Cognition has developed and patented. Retail, in other words, is a single case study for the power of this technology. Even so, it’s proving to be a tremendously lucrative one: Suswal told Bangash that the company has inbound interest from more than 300 retailers, including some of the world’s largest retail chains. Standard Cognition is already “drowning in demand” for its core service, the retrofitting of store locations with cameras and their SaaS technology. This summer, the company announced a major partnership in this arena with Alimentation Couche-Tard, the Canadian parent company of several retailers, including Circle K convenience stores. Standard Cognition will begin retrofitting 30 Phoenix-area Circle K stores with its computer vision-enabled cashier-less checkout system. If the pilot is a success, the technology will deploy to the rest of the chain, and perhaps to some of the other stores owned by Alimentation Couche-Tard, the fifth-largest convenience store operator in the world.
If none of this sounds terribly futuristic, consider some of the other potential applications of the technology just within the retail sector. In the first place, average retail stores experience a 2% shrink in inventory due to damage and theft. And on average, large retailers spend about 2.5% of their budgets on anti-shrink measures, such as store security systems and human guards. This, Suswal signaled, is already a meaningful market opportunity for Standard Cognition. But as Suswal also points out, the data generated by the Standard Cognition system will likely prove even more valuable.
The Standard Cognition system accounts for a product’s entire journey through the store: At what point does a customer pick up a product? When do they decide to put something back? And what products does a certain type of customer pick up, consider, and put back down? These are the sorts of questions that Standard Cognition data will enable enterprise CPG (consumer-packaged-goods) companies to answer, inspiring them to change the shape, design, or packaging of their wares in response. For retailers, the data will provide solutions for inventory management. Retailers are also asking Standard Cognition to create operations solutions for employee management, such as logging employees’ time at work without relying on them to clock in and out.
Particularly among larger retailers, such as supermarkets, Standard Cognition will free up employees formerly dedicated to checkout lines, and enable merchants to redeploy workers back to human-centered tasks, such as assisting customers on the floor. Suswal describes the changes enabled by the technology as the “new face of retail.”
A Future Financial Juggernaut
Because of its numerous potential applications, Suswal insists that Standard Cognition not be confused with its competitors in autonomous checkout: Amazon Go, for example, builds new stores; it doesn’t retrofit them. More importantly, the Amazon technology for autonomous checkout is entirely different: it relies on shelves fitted with sensors that complement the cameras on the store ceiling. In this system, identification of a product occurs by understanding where it came from on a shelf, not what it looks like. Standard Cognition, by contrast, operates only with overhead cameras that can identify individual items no matter where they are in the store, even if they’ve been misplaced by another customer. The Standard Cognition installation and integration can take place without remodeling the store to accommodate the new shelves, refrigeration units, and so forth. According to Seswal, the result is about 10 times cheaper than what competitors are able to offer retailers. The technology is also applicable to warehouse logistics and shipping, among other industries.
Becoming the leading company for retrofitting brick-and-mortar shops to autonomous checkout would already be a coup. But the opportunity set is much larger when viewed within a larger panorama. Designed to be compatible for any store using its technology, the Standard Cognition app could soon become one of the largest payment processors in use, making the company a rival to Square and Visa. If Standard Cognition succeeds, we believe they will be creating a trillion-dollar opportunity.
The Opportunity Set for LPs
Large oversubscribed venture funds have been in the SaaS game for more than a decade, and many have already cashed in on some of the sector’s biggest exits. Numerous promising SaaS companies remain privately held and venture-backed, but many are already at Series C and beyond. SaaS-enabled deep-tech and computer vision startups like Standard Cognition, however, are only just getting started, and offer LPs the opportunity to gain exposure in an area that we believe can yield strong returns as computer vision is applied to a broadening array of use cases.
Consider Ouster, another company working in computer vision. Ouster makes digital lidar sensors with automotive, robotics, and smart infrastructure applications. Ouster’s core product is currently priced at $600, but the company’s mission is to bring digital lidar sensors to an affordable $100 price point—an offering that could shape not only the future of autonomous vehicles, but of entire infrastructure sub sectors, including warehousing, construction, infrastructure development and retrofitting, and manufacturing. On December 22, Ouster announced it would go public in a reverse-acquisition with Colonnade Acquisition Corp., a special-purpose acquisition company (SPAC), in a deal that values Ouster at $1.9 billion. Ouster is the fifth lidar manufacturer to announce it would go public in 2020.
Indeed, the opportunity set is large enough to generate funds dedicated to seeding computer vision companies, such as LDV Capital. Established in 2012, LDV funds companies at the seed and pre-seed stages across Europe and North America. Its portfolio offers proof of the ever-widening range of applications for computer vision, which extends beyond transportation and logistics to agriculture, health care, smart home technologies, satellite imaging and mapping, and more. In the $8 trillion global agricultural sectors alone, computer vision can enhance efficiencies with such diverse tasks as phenotyping plants, predicting yields, and monitoring livestock on the production side, and with quality assurance, supply tracing, and food safety on the distribution side.
LDV estimates that by 2022, there will be 45 billion cameras in the world, collecting massive amounts of data. This web of cameras and data has created what LDV founder Evan Nisselsen calls the “Internet of Eyes,” a network powered by the deep-tech applications and services being developed by its portfolio companies. In each of the sectors where computer vision technologies are deployed, they will do more than capture and transmit visual data; they will also teach deep-tech algorithms to understand the needs of humans and human-built systems and to solve complex problems.
As cloud computing becomes more accessible, startups offering physical hardware backed by powerful cloud-based SaaS systems for intelligent implementation are proliferating. We feel that LPs looking to gain exposure in new deeptech markets would do well to begin diligencing managers in these sectors. While autonomous vehicles and cashier-less checkout steal headlines for their easily imagined disruptions, the applications for these technologies stretch across entire industries in manufacturing, retail, and logistics—and across the entire globe.