Shaoshan Liu, a AAAS Member working in artificial intelligence, says his obsession with robots began with “The Transformers,” the popular 1980s American cartoon series that originated from its own action figures. The show spawned video games and several movies.
Liu grew up just north of Hong Kong in Guangzhou, China, so the show did not make its way to him until the 1990s, when he was 10 years old. The cartoon centers around the Autobots and Decepticons, two alien robot groups that are in constant battle. They “transform” themselves into other forms, including vehicles and animals.
Of the fictional robots, Liu says there is some truth. “You can see that robots can do a lot,” Liu says. “They can do harm, but they can also do a lot of good.”
Once in high school, Liu started building small robots out of Lego kits. He began with the sets that had motors he could not program, but eventually learned to build motor robots and would tinker with them to make the machines more autonomous. Liu even programmed a particular robot to travel 20 feet from his bedroom to the living room.
Liu’s career has come full circle. He is now the founder and chief executive of PerceptIn, a startup company that produces autonomous vehicles using computer vision-based technologies that could make them more affordable. He is also a researcher for Autonomous Machine Computing systems, an intelligent robotics company. Liu says his purpose in life is improving human society through technological advancements. To that end, his research centers around advancing people’s understanding of autonomous machine workloads.
One such project focuses on autonomous mobile clinics, self-driving vehicles with medical diagnostic equipment, telemedicine capability and artificial intelligence software that can perform health professionals’ primary tasks, such as disease screening and basic diagnostics. These clinics have the potential to revolutionize healthcare delivery by bringing these services to populations in hard-to-reach areas.
“I believe that this integration of technologies will revolutionize healthcare delivery, thus solving both the equity and access problem commonly encountered in the healthcare system today,” Liu says.
As part of this project, Liu is developing, open sourcing and standardizing health AI technologies to improve healthcare access in developing countries. In response to the global monkeypox outbreak, he is also working on an AI-based monkeypox detector called AICOM-MP that targets mobile devices and embedded systems so people in developing countries can more readily utilize advanced technology for public health purposes.
Liu also spends his time discussing the need for artificial intelligence regulations to have global standards. The problem, he says, is that countries have different regulatory standards, which favors large corporations because they have amassed massive legal departments to focus on these standards. Liu estimates his start up PerceptIn spends 40% of their budget ensuring they are in compliance. In his view, the lack of global AI regulation standards impedes the integration of AI technologies into our society.
“When we deal with these different governments, they are strict … so sometimes, they overregulate,” Liu says. “For AI, there’s no equivalent of the FDA in the United States. For AI, there’s no golden standard to follow.”
The autonomous vehicles his company makes cost consumers between $10,000 and $20,000, and Liu keeps the price tag relatively affordable to cover trips under five miles. His research found that these trips, also called micro-mobility, are underserved due to their high cost with non-autonomous cars. While young and able-bodied people in cities might walk, rent a scooter or bicycle, people with disabilities and aging populations do not necessarily have those options.
Research from the U.S. Department of Transportation found that in 2021, 52% of all daily trips taken in the U.S. were less than three miles and 28% of trips were less than one mile.
“Autonomous driving, if you can make it more affordable for folks on those short trips, will bring a lot of benefits.”
PerceptIn has begun delivering autonomous vehicles for micro-mobility services to various towns where elderly people have difficulty accessing public transportation, areas near industrial parks and various tourist sites in both Japan and China.
Americans can eventually expect to see widespread use of autonomous vehicles on the roads, although Liu states that will take a while.
In the past decade, there has been tremendous growth in the global intelligent electric vehicle (EV) market, but many of these vehicles’ original equipment manufacturers compete on their autonomous driving technologies. That has resulted in accidents, he says.
In July, the National Highway Traffic Safety Administration opened a probe into a crash involving a Tesla Model 3 in California that killed a pedestrian. In August, China’s XPeng P7 that was traveling on a highway at a high speed rear-ended a parked vehicle it failed to detect, killing the person.
“As a result of all the fatal accidents that involve intelligent EVs, it leads me to believe that we should revisit how we ensure the safety of intelligent EVs powered by advanced autonomous driving technologies,” Liu states.
Pointing to technical and regulatory concerns, Liu suspects it will take years before Americans see large-scale commercial deployment of autonomous vehicles. And that is not necessarily a bad thing.
“Intelligence, after all, should be elegant instead of ruthless,” Liu believes. “I think now it is time for the industry to step back a bit from the technology race and reach a consensus on delivering the safest autonomous driving technologies for customers."