Living on the Frontier

From Smartphone Guts to the Future of Compute

Author: Patrick Eggen (@peggen), Qualcomm Ventures Senior Director, Early Stage Fund

** UPDATE (3.11.16): The automotive/robotics investment noted below is Cruise Automation, recently acquired by GM for $1BN+.

As early-stage lead for Qualcomm Ventures, the investment arm of Qualcomm Incorporated, I’m constantly meeting with companies that reflect the next generation of compute platforms beyond the modern-day smartphone. As an investment group, we’ve been inspired by the oft-referenced (and rightfully so) 2013 phrase from Chris Anderson: “peace dividend of the smartphone war” which serves as an elegant metaphor for the hardware renaissance. Anderson posits that the guts of a smartphone — consider all the sensors, GPS, camera, core processors, memory and so on — are now available today for modest prices. This has enabled start-ups to build viable low-cost devices. I’d take this position a step further and say that this trend has effectively created the entire early-stage hardware and connected device ecosystem of today. The smartphone is effectively catalyzing all these new compute platforms.

My team identified this trend early on and actively deployed capital along this framework. As part of a world leader in mobile technologies, we didn’t want to sit on the sidelines. We wanted to make bold bets irrespective of the technological and regulatory unknowns. We wanted to ride the peace dividend theme into the frontier.

Given this, Qualcomm Ventures (QCV) has emerged as a leader in early stage “Frontier Tech” — a term coined in recent CB Insights reporting. Per the following visual, QCV has made significant investment in the space:

Specifically, we’ve concentrated our early stage capital across 3 different themes:

(1) Sensor and machine learning enabled use-cases that leverage existing smartphones (i.e. crowdsourced data, business insights)

(2) Next-generation multimedia based on ever growing compute capabilities of high-end smartphone (i.e. AR/VR/3D data capture)

(3) And, most importantly, potential next-gen compute platforms and native experiences that evolve from mobile (i.e. drones/automotive/robotics)

Below provides a sense of our current focus in these select, disruptive categories:

AR/VR

While many have been enamored by the category, few investors have the appetite to invest today. QCV was a pioneer in this space and identified category leaders at the early stage. On the augmented reality front, we scoured the world for companies that had both technical chops (computer vision-based) and Madison Ave marketing pedigree. For instance, we were the sole Seed and Series A investors in UK-based Blippar which is a unique AR company that’s monetizing at scale. Recently, they launched their visual search platform in an ambitious attempt to index the real world. QCV was also an early investor in Magic Leap, the leader in next generation augmented reality creating the category known as “cinematic reality”. We also invested at the Seed round in Matterport, which offers a comprehensive, full-stack 3D data capture platform. Matterport’s technologies let users automatically create immersive 3D digital models of indoor spaces. Once 3D sensors become ubiquitous in the smartphone, we believe that this will unlock a number of consumer use-cases. Additionally, Matterport resolves the “content creation bottleneck” in virtual reality as it’s the rare company creating immediately digestible content for this new medium.

Artificial Intelligence

The subject of artificial intelligence is everywhere these days, but how close are we to substantial impact? To date many applications of AI have come off as arguably underwhelming. Enter deep learning based on neural networks.

Deep learning involves effectively “training” a computational model to deduce meaning from incoming information. Our portfolio company Clarifai uses this approach to bring the world’s best image recognition technology to market. Local image recognition in the form of brand logos, auto tagging and even people’s faces presents overwhelming opportunity for future use-cases and business models, particularly as it relates to mobile. As its software can rapidly analyze video clips to recognize over 10,000 different objects or types of scene, we believe that this will change the way you search for photos or videos on your phone.

Drones

QCV has quietly amassed one of the more prolific drone portfolios in the market, having invested in 6 companies in the space — fitting since this “smartphone with propellers” represents another emerging compute paradigm. I work closely with my colleague, Gareth Keane (@garethekeane), who drives most of our investing activity here. We led the last round of 3DR, which takes a full stack approach and produces ready-to-fly consumer drones (think sub-$1000 at your local retail store). In terms of enterprise use-cases, QCV was an early investor in Skycatch, which offers the most comprehensive end-to-end solution aerial data imagery related to drones. We’ve also backed enablers in the drone ecosystem; namely Swift Navigation, which offers precision GPS solutions down to the size of a nickel. This is critical for UAV enterprise use-cases in precision agriculture, automotive and mining.

Robotics

Robotics further leverages all the components of the smartphone revolution. Think of it as a “smartphone without the screen”. We believe that robots could fundamentally change the nature and use of human labor across industries in the future. More specifically, the next wave of marketplace innovation could see the introduction of robotics to create much more attractive unit economics.

Varun Jain (@jvarun83) from the QCV team, highlighted two of his portfolio companies that are exploiting this disruptive trend in a recent post. While both investments are not public yet**, the first one focuses on the urban transportation market by building an on-demand self-driving taxi service for congested city centers. The second targets the food/package delivery market by building an on-demand drone delivery service for suburbs. While they attack different markets, both share the same disruptive theme of offering superior services at a lower price point than the legacy human-powered marketplaces. While still in the early days, I see a similar analogue playing out in the food technology space as evidenced by our North American QPrize™ winner, teaBot — a full-stack tea-vending robot.

Automotive

The automobile represents another frontier compute platform. Historically this giant, sleepy industry has been saddled by slow product cycles that are now being fundamentally disrupted. We are spending our time in 4 primary areas: (1) telemetry; (2) infotainment; (3) advanced driver assistance systems (ADAS); (4) autonomous vehicles. The combination of low-cost breakthrough sensors (i.e. LIDARs) and sensor fusion algorithms will bring advancements like lane detection, collision avoidance and drowsiness detection to the vehicle. QCV is an active investor in the category with exposure in the infotainment and autonomy buckets. We have hedged our bets in this market by investing in solving the biggest problem for the 250M existing cars on the road via heads up display provider Navdy, and at the same time investing in the aforementioned self-driving car company**, which can fundamentally replace human drivers. Given the regulatory unknowns and market timing issue, this justifies our hedged bet.

Going Forward

New categories necessitate big, early bets — bets that we’re willing to take. As experts in mobile, we are especially comfortable backing new applications and compute platforms which leverage the smartphone. With hardware technology as plentiful and cost efficient as it is these days, a new wave of IoT disruption is just around the corner — and these connected devices will only get smarter and more useful with each new sensor that is added. Akin to how Waze exploited GPS to spawn a new era of location aware apps, each new sensor makes a new generation of applications possible. This is the underlying thesis behind our OpenSignal and Sense360 investments.

By using our knowledge of the mobile space as a guiding light, we see huge promise in (1) Sensor and machine learning enabled use-cases that leverage existing smartphones; (2) Next-generation multimedia based on ever growing compute capabilities; and (3) The ongoing evolution of the mobile compute platform into new verticals.

In 2016, we look to continue to push what constitutes the frontier and identify new categories and opportunities beyond our current comfort zone.

“Where the frontier of science once was is now the centre.” ― Georg Christoph Lichtenberg