MODE, which provides a wide range of on-site IoT platforms for various industries, has announced a business partnership with Canon Marketing Japan (Canon MJ) and Safie. By combining sensor data with video data and generative AI technology, the two companies will develop more advanced business support solutions.
MODE, which provides an IoT platform for work sites in a wide range of industries, including construction, manufacturing, and retail distribution, announced a business partnership with Canon Marketing Japan (Canon MJ) and Safie on February 12, 2025. By combining sensor data with video data and generative AI technology, the company will develop more advanced business support solutions. A generative AI assistant evolves the IoT data platform for work sites BizStack, which MODE will provide from 2022, is an IoT platform that collects, integrates, and manages sensor data (IoT data) from all work sites. In 2023, the company will also add the “BizStack Assistant” function using generative AI technology, allowing the AI assistant to answer questions based on sensor data via business chat (Microsoft Teams, direct, etc.). MODE co-founder and CEO Manabu Ueda said that last year (2024), the BizStack Assistant was introduced and started to be used at large construction sites and other places. ” the AI assistant resides on the chat system, and users can give instructions and ask the AI assistant questions in the same way as they would talk to other members. In the future, the AI assistant will also participate in real-time voice communication via intercoms and walkie-talkies, and it will be possible to give voice instructions and receive voice responses and reports,” (Mode’s Ueda) As the use of BizStack evolves in various work sites, a new need has emerged: the advanced use of “video data.” For example, even if BizStack detects changes in sensor data and issues an alert, humans still want to see what is actually going on at the site. For this reason, “we found that it would be very useful if there was a system that allowed users to smoothly check the changes captured by the sensor through video.”
Another evolution is the emergence of “multimodal LLM” in the field of generative AI. Multimodal LLM can understand not only the contents of text data but also the contents of images/videos. By incorporating this technology, it is possible to use it to support various on-site work. As concrete examples, MODE’s product manager, Hiouma Watanabe, introduced simple examples of use, such as “notifying if someone without a helmet enters the work site,” “displaying a graph of the on-site congestion status as time-series data,” and “counting how many ready-mix concrete trucks (concrete mixer trucks) entered the work site yesterday.” He said that the plan is to gradually implement such functions using video data as additional functions (skills) for BizStack Assistant. Ueda said that they would like to introduce simple functions such as calling up video via the AI assistant within a few months, and functions such as issuing monitoring instructions to the AI assistant to perform continuous video monitoring by the end of this year. The two video solution companies also have high expectations for the business partnership.
こちらもお読みください： 第3のAIソリューション、マルチAIエージェントシステムの接続を拡大
MODE has previously partnered with more than 50 IoT sensor device manufacturers through its partner program, enabling collaboration with the BizStack platform. This time, the company will partner with Canon MJ and Safie as partners in the camera video field. Safie boasts a high domestic market share in the cloud service field for monitoring and recording camera video. Kazuma Morimoto, head of the company’s development department and CTO, said that the company’s services are widely used across a range of industries, and the company currently has 27.50,000 units (as of the end of September 2024). Morimoto said that he expects two things from the partnership with MODE: expanding the scope of problem-solving through collaboration with various sensor data and improving the value of the video data itself, and combining it with generative AI to enable flexible and diverse control and realize advanced image analysis. Safee is working on building an “AI solution platform” that will streamline the development of AI applications using camera video data, but said that in the future it would also like to consider linking sensor data via BizStack. Meanwhile, Canon MJ is focusing on the video solutions business as one of its focus areas in the IT solutions business. As it promotes its business, the positioning of camera images is “moving from the phase of simply ‘watching/taking’ images to the phase of ‘utilizing’ them,” explains Terakubo Tomoaki, director of the company’s Solution Development Center. “Customers who have entered the utilization phase are saying things like, ‘It would be even more convenient if we could connect this video to other systems,’ ‘It would be even more convenient if we could see the video of the site, not just the numerical information from the POS and sensors,’ and ‘It would be even more convenient if the procedure manual needed for the next task could be displayed immediately after viewing the video.’ In other words, they are placing importance on how to make the workflow as efficient as possible,” said Terakubo of Canon MJ. To meet such customer needs, he explained that a solution is required that allows video data, sensor data, and even digital documents such as manuals to be easily and collectively viewed on-site. With this in mind, he was shocked when he saw MODE’s BizStack Assistant, and said, “I thought this was exactly it,” explaining the background to the decision to partner. Since the 2010s, the Canon MJ Group has continued to acquire companies and form capital and business alliances in fields such as network cameras (IP cameras), VMS (video management software), VSaaS (video SaaS), and video analysis software/AI. Through this partnership with MODE, which has a platform and front-end application, “these things are starting to organically connect,” he said, expressing his hope that the areas in which problems can be solved will expand.
Meanwhile, Ueda of MODE cited two reasons why モード chose these two companies as partners: the imaging solutions of both companies have already been introduced in many sites, and when they saw MODE’s technology, they quickly understood the value, and negotiations for the partnership proceeded smoothly. キヤノンMJ and Safee have already been partnering in both capital and technology since 2017, and are working on joint solution development and cross-selling. MODE’s Ueda described it as “a form of participating in that (framework).” On the same day, MODE also announced that it had raised funds for a Series B additional round. The company raised a total of about 800 million yen (5.3 million dollars) from four companies: Canon MJ, Safi (Safi Ventures), KDDI Open Innovation Fund, and Kirin Health Innovation Fund. The total amount raised in Series B is about 2.1 billion yen, and the total amount raised is about 4.57 billion yen (29.76 million dollars).
ソース ヤフー
