Mar. 16, 2022
Tadano Sees Strong Results From Its First Open AI Innovation Competition
The Tadano Core Values are Safety, Quality, and Efficiency based on Compliance. For safe and efficient crane operations, trained operators who can make smart, fast, and accurate decisions are ideal. However, challenges such as the declining and aging workforce make it difficult for companies to find enough operators. Technology shows great promise to help solve this problem.
To that effect, Tadano recently organized its first full-scale open innovation competition with a goal of creating a motion control AI that can execute crane operation simulations.
Mr. Fumihiko Futaba, Assistant Manager of the Tadano Advanced Research Technology Center, oversaw the competition. He noted that while Tadano has already engaged in joint research with companies and universities, this public competition offered an opportunity to “take open innovation activities one step further.”
Tadano partnered with SIGNATE, a data science platform that promotes problem-solving collaboration efforts between companies, government agencies, and research institutions, often through challenge competitions. Tadano’s crane simulation challenge proved one of the most challenging and rewarding, as it offered the highest prize money in their competition history.
1,033 participants accepted the challenge and a total of 1,150 entries were received. On January 28, 2022, an online ceremony was held to award the winners, who gave presentations of their individual or team solutions, followed by a Q&A session and commentaries from experts.
“We are very grateful for the number of applications we received,” said Mr. Futaba. “This competition was not a simple data analysis, but a difficult and unprecedented task of running a simulator to build a control AI. The fact that many people participated is proof that there are many online researchers and engineers who are full of inquisitiveness, and we felt the possibilities from such open innovation.”
“Tadano makes heavy equipment products that are large and sophisticated. These cranes operate in challenging environments, which are impacted by such trends as decarbonization, the aging of the general operator population, and lifting operations characterized by long wait periods with suspended loads,” said Mr. Toshiaki Ujiie, President and CEO of Tadano Ltd. “To meet these challenges, it is necessary to take on such product development initiatives as electrification, automation, and remote control.”
The complexity of this challenge required seminars with technical support to educate participants on cranes and their operation. Participants were judged on the speed, accuracy, and safety of load lifts executed during arbitrarily determined slewing operations, all of which require complex calculations.
“Crane operation requires skill; it is difficult to balance out misalignments, while managing load sway and speed,” said Kaoru Nasuno, CEO of DeepX, Inc., which provided the simulator. “With this in mind, the winners' innovative approaches were very impressive. However, when those algorithms are applied it is necessary to adjust details such as dealing with load sway and feedback control against wind.”
One participant discussed the importance of Tadano’s challenge to make a true impact. “The use of AI is often talked about, but there are not many cases that essentially enrich people's lives. In this sense, the automatic control of the crane was very meaningful. Also, there was room to devise various approaches.”
Another participant noted: “I think the lack of engineers with reinforcement learning expertise is a barrier we must break through. It is very significant that Tadano took the initiative; it may become an exemplary case not only for manufacturers of heavy equipment such as crane such as cranes but also for various industries.”
After careful analysis and judging, winners and other finalists were invited to the awards ceremony, held on January 28th. Around 80 people were invited to watch the top three finalists giving presentations on their solutions with technical details. A lively question-and-answer session followed, often revealing ingenious development methods.
Among the five overall winners, one was a team and four were individuals. The top winner was Team Nssol, which consisted of four members of the Systems Research and Development Center of NS Solutions Corporation. Team members researched and devised solutions using a variety of methods and embraced the opportunity to improve their knowledge of reinforcement learning. “Through this competition, we got to work with cranes, which are products we would not normally deal with. Also, the format of competition, by a provided simulator, is also rarely seen anywhere else. For us, the competition was very interesting and the experience very enjoyable. Thank you for providing such an interesting theme. ”
“Reinforcement learning technology is advancing day by day, and ingenuity is required to evolve,” said Professor Yutaka Matsuo, a leading AI researcher at Tokyo University and chairman of the Japan Deep Learning Association. “I think that each team realized that they could do something good by devising solutions. With this as an opportunity, I hope more will take on the challenge of using AI for machine control.”
Participants certainly learned a lot about cranes during the challenge, and some found greater appreciation for cranes: “When the competition was over,” one commented, “I was a little attached to cranes, and it became fun to see a construction site.”
These results have proved encouraging for Tadano’s AI efforts and new collaborations. Mr. Futaba said, “Tadano is researching and developing AI adaptation to other products, and there are some fields where promising results are being produced. In the near future, we would like to set up other channels to disclose our AI efforts, so please look forward to our future developments.”
Mr. Ujiie added, “Cooperation with external partners is indispensable for Tadano as we take on new initiatives. I was very impressed by the results from this competition. In a manner that could not have been achieved by our engineers working alone, the participants produced wonderful solutions in a very short period of time. Tadano will continue to pursue such open innovation opportunities proactively.”
