You’ve heard about autonomous cars, but what about autonomous advertisements? Meet PerceptIn‘s DragonFly Intelligent Advertising Vehicle, a driverless chassis equipped with digital signage — and a vending machine to boot.
The mini-shuttle sports four displays — two smaller screens on the front and rear and a pair of widescreen monitors on either side — plus a scrolling message board up top. It prominently features a tactile button that activates the aforementioned vending machine, which spits out the beverage or other merchandise of the operators’ choosing. And thanks to a bevy of vision sensors and an AI-infused software platform, it can be hailed with arm gestures from passersby.
The Intelligent Advertising Vehicle is the brainchild of Dr. Shaoshan Liu, a University of California Irving graduate and former Baidu engineer. He’s the founder and CEO of PerceptIn, a robotics startup headquartered in Santa Clara, California that’s been developing an in-house self-driving car platform since 2016.
“There’s no good monetization model for autonomous driving yet,” Liu told VentureBeat in a phone interview. “When vending machine manufacturers came to us and asked, ‘Hey, can you make a vending machine move?’ We asked what the benefit was, and as it turns out, the movement attracts a lot of attention from people. The goal is to convert that attention into buying activities and advertisement viewing activities.”
PerceptIn has deployed a number of prototypes in an industrial park campus in Shenzhen, China, built on top of its modular self-driving tech. They’re powered by Nvidia’s Jetson platform and equipped with four high-definition RGB cameras, radars, ultrasound sensors, and GPS, in addition to odometry sensors that count the number of rotations completed by each wheel. With the odometry data alone, the company can pinpoint the Intelligent Advertising Vehicle’s location to within 20 centimeters, and adding cameras to the mix enables them to construct a panoramic point cloud at 30 frames per second.
Said cameras can also perform rudimentary face tracking and gaze detection, which advertisers can use to estimate engagement.
“We’ve calculated [what it takes to get] people to actually pay attention to an ad,” Liu said. “When people spend more than five seconds staring at an advertisement, the [engagement rate’s] crazy — it’s 58 percent.”
Currently, PerceptIn isn’t selling the Intelligent Advertising Vehicles — it has a fleet of about ten that it rents out for conferences, conventions, and one-off promotions for $3,000 per day per vehicle. But the company is considering a “regular operation” model where it would supply the driverless machines to office parks, universities, amusement parks, or even cities for a one-time fee.
As part of a pilot program in the U.S. that’s scheduled to kick off in early 2019, customers will be able to purchase an Intelligent Advertising Vehicle for $40,000.
Self-driving ad billboards — and vending machines, for that matter — aren’t as wild as they might sound. A Forrester report last year — “Autonomous Vehicles Will Reshape the Global Economy” — speculated that driverless cars could someday become glorified mobile theaters. And at the STEP conference in Dubai this year, Pepsi piloted a driverless cart that allowed customers to buy drinks through Facebook Messenger.
Their practicality might be up for debate, but they certainly draw a crowd. Liu didn’t have hard numbers to share, but said that anecdotally, at busy events, the Intelligent Advertising Vehicles often attracts “hundreds” of people within 15 or so minutes.
The Intelligent Advertising Vehicles’ debut comes after the launch of PerceptIn’s DragonFly Pod, an autonomous vehicle priced at just $40,000. Uniquely, it’s fully modular, allowing customers to swap out components on the fly.
“People often say that building an autonomous vehicle is just as difficult and far-fetched as building a rocket, but we disagree,” Liu said in an earlier interview. “[We’ve] simplified the whole process to make the building of autonomous vehicles not that much different from that of building with Legos … We are committed to the task of enabling people to create AVs and not just use them.”
PerceptIn emerged from stealth in March, but it’s already raised a decent chunk of change: $11 million overall, including a $2 million seed round in 2016 and an $8 million series A in 2017. Samsung Ventures, Samsung’s investment arm, was an early partner, and PerceptIn said it has “visual intelligence” hardware and software projects in the works for over 100 customers, including Huawei, ZTE, UC Irvine, Wayne State University, a partner in Detroit, and unnamed industrial parks in China.