LAKE WALES, Fla. — The $5 million IBM Watson AI XPrize competition, which kicked off last year and will end in 2020, was the first of the XPrize contests (14 since 1995) to have a contestant-defined “open” goal rather than a predetermined objective. Now it is also the first XPrize to add a wild card, giving new contestants until Dec. 1 to join the 147 teams that made the first-year cut.
“The total number of teams officially registered stands at 147, out of the total of 870 team submissions that were recorded from more than 9,000 initial interested requests,” Amir Banifatemi, prize lead for the IBM Watson AI XPrize, told EE Times. “Given the rapid pace of artificial-intelligence breakthroughs and the possibilities that AI opens to solve grand challenges, we wanted to ensure that teams with new ideas still had the opportunity to participate.”
The original teams still in the competition hail from 22 countries in total: Australia, Barbados, Canada, China, the Czech Republic, Ecuador, France, Germany, Hungary, India, Israel, Italy, Japan, the Netherlands, Norway, Poland, Romania, Spain, Switzerland, the United Kingdom, the United States, and Vietnam. Their projects are being evaluated not only for their efficacy in addressing AI challenges but also for their potential social, ethical, and technological impact. Imbuing AI with the ability to understand human emotional cues, for example, could have implications beyond the AI’s cognitive-computing capabilities.
Abel’s “eye” will use Watson to perceive, integrate, understand, and analyze data at levels beyond human abilities for edge-collaborative cloud robotics, according to the U.S. team.
The IBM Watson AI XPrize is also the first XPrize to leave the goal open to the contestants’ discretion, and as a result the teams have proposed ideas for solving problems across a wide swath of disciplines. “Energy Efficiency” projects would reduce greenhouses gases and makes landfills smart at separating recyclables. “Health and Wellness” investigations look to head off mental health problems, diagnose an infant’s crying, and improve sleep. “Learning and Human Potential” projects aim to reinvent computer coding, personalized learning, peer-to-peer tutoring, and scalable learning to achieve universal worldwide literacy. Proposals for “Improving Society” would automatically flag “fake news” in social media and get legal information to victims at little cost. “Shelter and Infrastructure” projects aim to meld social development with satellite imagery, predict disasters, manage traffic flows in cities, and assess the structural health of buildings. “Space and New Frontiers” explorations seek to develop neurologically inspired models and automatically propose hypotheses.
“We have been impressed with the level of variety and domain focus so far. Teams are very diverse, [hailing from] startups, academia, large corporations, nonprofits, and more,” Banifatemi said. Among the “impressive” entries, he said, are AI applications to “detect crop disease in Ethiopia, detect illegal mining in Congo, model malaria-prone regions in India, predict psychiatric medicine effectiveness, automate project management at scale, [advance] triage emergency medicine, and monitor the structural health of buildings.”
The RYTHM team, from France, is working on restful-sleep solutions that monitor, analyze, and stimulate your brain to enhance your sleep experience.
The addition of the wild-card teams aims to widen the application domains even further, but the expanded pool will still be subject to the same annual culling process. “Each year, up to 50% of teams will move to the next round provided they reach their milestones and are selected by judges to move forward,” Banifatemi said. “Based on how many wild cards are approved to compete, we expect to have half of the total teams in competition by September 2018 moving into 2019.”