The Age of Robotization is Upon Us
Growing Number of Products Will Have Perception, Insight, Intelligence
By Shaoshan Liu and Zhe Zhang, PerceptIn, Co-Founders
In 1969, in The Graduate, Mr. McGuire told Dustin Hoffman’s character “just one word: Plastics.” In 1982, Microsoft’s Bill Gates might have said “software.” Several years ago, venture capitalist Marc Andreessen took it a little further, and said that “every company is a software company,” a nod to the fact that digital disruption was a competitive advantage in most enterprises.
Flash forward to 2017 and beyond. We believe every product will be a robot, or at least have a robotic element to it. We call it robotization, the act of integrating robotic features into previously “dumb” inanimate products. Look at where we are today. It’s commonly accepted that basic service machines — vacuums and many of the machines used in various manufacturing processes — are robots. The pace of this robotization has increased quickly as innovative companies race to provide solutions for indoor tracking, mapping, VR/AR tracking, and even early stage unmanned vehicles. Commonly referred to as self-driving or autonomous cars, these vehicles are really just big, four-wheeled robots.
This shouldn’t come as a surprise. In 2007 Bill Gates wrote an article entitled “A Robot in Every Home,” in which in envisioned a future in which robotic devices would become a nearly ubiquitous part of our day-to-day lives. There’s plenty of hype in the technology business — but this prediction is very well on its way to becoming reality.
Everyone In the (Robotics) Pool
A growing number of companies are working on this. The list includes major Silicon Valley firms, including Intel and Google, as well as car manufacturers and innovative startups such as our firm, PerceptIn. Like most quickly evolving markets, it’s fragmented as we all collaborate, partner and compete to create the best path forward. The market is large, with almost limitless robotic applications, so there will be many winners and many companies required in order for this eco-system to thrive. We believe we’ve just scratched the surface of Visual Intelligence’s capabilities, we can expect this trend to be a major disruptor in the coming years; in fact, there’s no end in sight for this technology as it is set to transform nearly every sector of the economy.
Even what we consider existing, low-end robots can improve significantly. Household-based, indoor robotics products will get better at performing their own mapping and tracking within a home, and be truly autonomous, unlike current versions where LiDAR solutions work best on specific 2D planes. As these robotics products become more sophisticated, safety considerations and obstacle avoidance will be more important. At PerceptIn, our visual camera-based solution can map, track and perform obstacle avoidance. We’re working to overcome low-light conditions by fusing visual with other information (sonar, wheel, bumper, or even depth sensor), which will benefit next-generation outdoor robots as well.
The advancement in robotization has its foundation in improved AI, computer vision and sensor technology, as well as reductions in cost; so much so where it’s now feasible to integrate full stack visual intelligence systems into just about anything. The major advancement in creating truly autonomous products has been the ability to give them “perception,” or the ability to understand the environment from sensing data, in a manner that is similar to the way humans use their senses to relate to the world around them. So you could say the new “plastics” or the new “software” is now “Perception Technology.”
This ability to improve a machine’s understanding of its environment is the basis for improving its reactions and enabling new forms of action. The industry doesn’t move forward without it.
Solving the Challenges of Perception, Performance, Power
At PerceptIn, we believe we’ve solved for the perception challenge via a low-cost, vision-based, multi-sensor, fusion-based perception system that considers and uses a combination of computer vision technologies, without relying on LiDAR. Part of this solution entails the ability to create high performance with limited computing power, especially in embedded systems. It’s made possible because of the proprietary algorithms developed to run on low-power devices which allow for the performance of tasks which usually can only run on mid-end or high-end computing resources.
So what are the possibilities once the high-performance, low-power challenge has been solved? At PerceptIn, we are working with a range of companies, from robot makers, to smartphone OEMs and chip makers, but are fielding inquiries from auto makers, service robot manufacturers and healthcare equipment manufacturers which want to disrupt their industries not with software, but with robotics. Applications typically fall into two types: functional and entertainment. Mobile gaming and AR applications are viewed by many smart phone manufacturers as the next big thing to differentiate themselves from their competition.
With all the different applications for robotizing products, we are experiencing the emergence of a variety of technology providers. It includes companies like PerceptIn, which will work with product makers and their engineers to customize the software and algorithms and provide the hardware and software development kit that allows applications to be built off this foundational technology. Others will merely provide the SDKs, and allow the product engineers to explore and create solutions on their own. We believe greater customized integration — such as well-calibrated and synchronized sensor module — is needed, along with collaboration to integrate the robotic vision into products to realize specific functions.
Applications As Far As the Eye Can See
A 2016 article in Nature said the robotics industry will grow from $25 billion in 2015 to $42 billion in 2020, and then jump to $77 billion by 2025. While it’s near impossible to quantify the number of robots we will encounter on a daily basis (at home and at work), our mission over the next five years is to robotically power — robotize — as many products as possible, from low-end IoT-grade robots to mid-range service robots to higher-end autonomous vehicles.
Although most of the talk of robotics and autonomous machines revolve around self-driving cars, we believe:
1.) Other types of autonomous products will outnumber and evolve more quickly than self-driving vehicles; and
2.) Autonomous vehicle technology has the potentially revolutionize the way we think about transportation; so much so that we envision a future in which traditional cars don’t exist.
We also see the future of cities being driver-less and believe we can help deliver the vehicles for a cost-effective price — at around $10,000 per vehicle. We know this would be seen as a radical step… but in a place like New York City, where CitiBikes have become super popular, and sections of the city are closed to traffic, the trend is to create a city less dependent on traditional manned vehicles.
There will be a time — perhaps in 15–20 years — where people will look back and chuckle when told that decades earlier people called their primitive phones “smart.” The advancement of several different technologies — including AI and computer vision — is quickly ushering in the age of robotization, and exploding what was formerly a small subsection of the technology industry. The burgeoning robotics industry — leveraging the combined efforts of established firms, along with innovative early-stage firms, such as PerceptIn — will be the force that affects the way we live more than any other industry.