Back to “the basics” in order to pursue new ideas

Connect human ideas to information through research.
With such a passion, Mr. Iku Ohama has been devoting himself to scientific research in AI as a researcher at the Business Innovation Division at Panasonic. He said, “Since Panasonic has been operating a variety of businesses, it covers a wide range of research fields.” His goal is to improve the efficiency of company’s core businesses. He also said “Scientific research plays the key role in the continuing development of AI, as the field grows very fast.”

His intuition brought him to AI research, which was not his area of expertise.

Mr. Ohama studied computer vision at his university. He entered Panasonic as a software development engineer for DIGA in 2006. After three years, he changed his research to AI. The idea to start something new spread through his division due to the sluggishness of the digital appliance boom at that time. He said “I was frustrated by being unable as a software development expert to work on unspecific phenomena such as human emotions in my work.” After getting to know AI research through an intern from a graduate school at the division next to him, he felt in his bones that he can create a device to read human emotions or bothers with this technology.

He asked his boss to work on AI research, saying “Please lead me through this project.”, at the time analyzing behavioral clustering for photographers based on the photos taken with their digital cameras. Although he didn't succeed in making a product, he devoted himself to the project and wrote a research paper. However, making a product is necessary for researchers at companies, different from those at other research institutions. He said with a smile, “There is always a boss who values scientific research and supports my unique plan.”

Research fields are “infinite”.

A large amount of data allows us to lower defect rates of manufactured products or to predict customers’ purchase behaviors.
One of the reasons Mr. Ohama, who works on AI research, decided to enter Panasonic is the wide range of its businesses. Research fields for AI technology are “infinite”. Therefore, he believes that skill to find out its potential field and talent to expand scientific research into related fields are required for an AI researcher.

Machine learning improves the efficiency of human decision-making and helps reduce labor force and cost.
He believes AI will become the next approach after “Kaizen”, which was created by The Toyota Motor Corporation and was applied in the manufacturing industry all over the world.
Development, management, and accounting.
Companies are actually operated by a lot of staff members with professional knowledge who are in charge of the most fundamental tasks. Based on the increase of the company’s operational efficiency by using AI, he believes that we can look ahead to spread the value of their research to the world by exporting it to other international companies. He thinks that this is the best advantage of working at a large manufacturing company.

Prepare for solving problems.

As AI has been rapidly adopted by us, Mr. Ohama has become interested in fully automated maintenance systems as well as in challenges facing AI. Although learning speed becomes faster by improving performance of a processor, the performance degradation caused by changes of its data quality is bothering AI researchers. He thinks that it will be difficult for AI to keep contributing to society without learning technology that supports maintenance tasks without manpower, including researchers and engineers, otherwise taking a lot of work. He said, “We haven’t succeeded in commercializing the best AI model that doesn't require any special adjustment once it has learned how to operate.” He believes that scientific research is the source to solve such problems.

He compares his research to a fire fighter, saying “Gaining stamina through daily workouts is essential for getting ready for the real thing.” He always tries to achieve something useful for others. Improving labor productivity through AI might play a key role in combatting the working population decrease that Japan is facing.