On December 15, 2021 (Wednesday), at Tokyo Port City Takeshiba, we will use Oki Electric Industry Co., Ltd.’s (hereafter, OKI) AI edge computer “AE2100” to compete for ideas and technologies for solving social issues. AIEDGE CONFERENCE & SOLUTION CONTEST 2021” was held.
The winner was Usagi Co., Ltd., which announced a tool that visualizes the human gaze, “Gaze Simulation AI”.
In this article, we will tell you about the state and results of the AI Edge Conference & Solution Contest 2021.
Table of Contents
- Part 1 AI Edge Conference
- Opening remarks by Mr. Tsuboi of OKI | Solving social issues through co-creation with partners
- Intel Suzuki, Keynote Speech | Four Superpowers and Intel’s Strategy
- Part 2 AI Edge Solution Contest
- Smart City
- Manufacturing
- Tevhnology
- Result announcement
- 3rd place Metro Co., Ltd. Prize money of 500,000 yen
- 2nd place Tokai Electronics Co., Ltd. Prize money 1 million yen
- 1st place Rabbit Co., Ltd. Prize money 2 million yen
- Ledge.ai Award (Special Award) Salty Star Inc.
- In conclusion
Part 1 AI Edge Conference
In the first part of the AI Edge Conference, “Opening remarks”, “Keynote speech”, and “Panel discussion” were held. People from tech companies active in
Opening remarks by Mr. Tsuboi of OKI | Solving social issues through co-creation with partners
Mr. Masashi Tsuboi of OKI, who made a speech at the opening of the contest, introduced the AI edge computer “AE2100” and talked about his thoughts on this contest.
Oki Electric Industry Co., Ltd. Director and Senior Managing Executive Officer, General Manager of Solution System Business Headquarters Mr. Tsuboi
OKI aims to solve social issues using an AI edge computer called AE2100, which was launched on the market in 2019. The AE2100 is an edge device that performs high-speed deep learning inference processing, and can perform AI analysis of large amounts of sensor data and video in real time and with high reliability.
Currently, we are starting to implement it in a wide range of markets and environments such as transportation, finance, manufacturing, and marine.
Mr. Tsuboi spoke of the importance of co-creation with partners, saying, “OKI alone cannot solve social issues (using the AE2100).”
| Mr. Tsuboi: When the idea and technology of AI edge came out, there was also a question as to whether it would really be useful to everyone. Therefore, we have created new things while conducting demonstration experiments as end users (companies) and partners for OKI. Currently, we see the entire DX ecosystem, including the 102 companies that support the AI Edge Partnership and collaboration with industry and academia. I am very grateful to be able to work with people from various fields. |
AI edge computer “AE2100” provided by OKI (Image source: OKI website )
The AE2100 is used in all the ideas of the nine companies participating in this contest. At the end of his remarks, Mr. Tsuboi said, “I expect this to be a very interesting contest where you can experience AI being used in various fields.”
Intel Suzuki, Keynote Speech | Four Superpowers and Intel’s Strategy
Mr. Suzuki, Representative Director, Intel Corporation
In 2021, the global semiconductor market was predicted to grow by 20% compared to the previous year, but the demand for personal computers and the cloud has increased due to the prolonged spread of the new coronavirus infection, and the actual growth is about 35 to 40%. It is said that. Currently, the problem is that the supply of semiconductors cannot keep up with demand.
Mr. Suzuki explained that Intel sees such industry movements as “four Superpowers.”
| Mr. Suzuki: We believe that the four fields of “ubiquitous computing,” “always-on connectivity,” “cloud-to-edge infrastructure,” and ” artificial intelligence ” will each become stronger, and their combination will lead to innovation. Until now, the pendulum has swung between “centralized” and “distributed” (*) in the world of computing. Currently, it is considered to be swinging to the “centralized type”. It is said that by 2025, 75% of all data will be generated outside the data center. Intel predicts that the pendulum will swing to “distributed” in February 2026, and predicts that the combination of AI and edge will grow significantly. |
Centralized type: Processing type in which most processing is executed on a central server in a one-to-many
manner.
Intel’s three strategies based on four Superpowers were introduced.
| product leadership | Providing not only GPUs but also XPUs (*), leading the democratization of computing (*) A computing architecture that can be used in a variety of applications, from PCs to cars and robots. |
| open platform | We will continue to invest in open platforms, especially in the software layer. |
| large scale manufacturing | We will invest heavily not only in our own semiconductors but also in our own manufacturing capabilities. |
Based on these, Mr. Suzuki said, “Intel will play the role of a neutral hub in the industrial world.”
Part 2 AI Edge Solution Contest
In the “AI Edge Solution Contest”, 9 companies that won the preliminary round (10 companies that passed the preliminary round, 1 company that declined) will compete for solutions using AI.
Solutions were divided into three categories: “Smart City”, “Manufacturing” and “Tevhnology”, and presentations including demonstrations were judged for each.
Smart City
| Intelligence Design Co., Ltd. | [Flow line analysis solution using person feature quantity matching] Automate the fact-finding survey, which was conventionally difficult. By linking the edge and the cloud, it analyzes a person’s feature values such as gender and age. A service that supports smart cities and is being considered for use in bus OD (traffic volume) surveys and surveys of actual visitor numbers. |
| Wants Co., Ltd. | [Predictive maintenance of utility supply equipment] By utilizing thermography and monitoring the temperature of important equipment such as generators and piping, signs of abnormalities can be detected. By performing AI analysis at the edge, detected abnormalities can be confirmed in real time, reducing the work cost of maintenance personnel. Used for predictive maintenance and quality control. |
| Computer Mind Co., Ltd. | [Analysis of clothes of indoor people by AI] A network camera is installed indoors to detect people and estimate their clothes in real time. It calculates a comfortable room temperature based on the number of people indoors and the clothes they wear, and automatically controls the air conditioning. Technology as an initiative for SDGs Goal 7, Goal 13, and No. 13. |
| Toa Radio Electric Co., Ltd. | [Monitoring smokers and littering in non-smoking areas] A service that alerts smokers using speakers, etc. when smokers are detected in non-smoking areas. Smoker monitoring video is analyzed in real time by an AI edge computer, and when a smoker is detected in a non-smoking area, a speaker or other alert is issued. It is also possible to notify only registrants with LINE WORKS reports such as smokers, smoking hours, etc. and real-time images. Technology that tackles SDGs Goal 3, Goal 11, Goal 12, and Goal 14. |
Manufacturing
| Salty Star Co., Ltd. | [Application platform & embedded hybrid database]”SpeeDBee Hive” is a service that contributes to solving problems at manufacturing sites. A wide variety of data can be analyzed. Data can be collected, analyzed and transmitted. The server function enables remote monitoring of data without programming by linking data via peripheral systems and graph tools. It is also possible to link with externally developed AI tools such as AWS and Azure. |
| Metro Co., Ltd. | [AI anomaly prediction by batch data conversion of multiple analog meters] A single camera analyzes the measured values of multiple meters in real time and notifies the results. It is possible to reduce labor costs and measurement errors by shifting from regular inspections by humans to regular inspections by machines. |
Tevhnology
| Usagi Co., Ltd. | [“Gaze simulation AI” that visualizes the human gaze]A line-of-sight simulation service that can visualize the line of sight of a person .By simply uploading an image, AI simulates and displays areas that are likely to be watched. Conventionally, eye tracking devices are costly and time consuming, but the gaze simulation AI can analyze images simply by uploading them, so costs can be greatly reduced. |
| Tokai Electronics Co., Ltd. | [Monitoring and preventive maintenance by visualizing odors] A service that detects and analyzes odors and humidity without contact using an odor sensor. It is possible to distinguish between urine and feces during excretion in diapers, based on how long the odor remains and the humidity. In nursing care, diapers can be changed quickly, which helps prevent rough skin and falls from bed. |
| BF Business Consulting Co., Ltd. | [AI coaching that utilizes swing sensors and posture estimation technology] A technology that aims to acquire a batting form as a super hitter that suits your body type. Clarifying the mechanism of athletes’ batting form and providing real-time AI coaching on the computer will improve the level of athletes. In the future, we plan to expand into other industries such as glass assembly for automobile manufacturers. |
Result announcement
The ranking of the competition was decided by the votes of the participants and the judges at the venue.
The 1st to 3rd place winners were announced, and a special prize was announced by sponsor Ledge.ai (Ledge Co., Ltd.).
3rd place Metro Co., Ltd. Prize money of 500,000 yen
Application solution: “AI anomaly prediction by collective data conversion of multiple analog meters”
| Mr. Tsuboi’s comment: It is a technology that embodies IT modernization (to modernize an outdated IT system by making it compatible with the latest technology), and it was highly evaluated for its potential to be very effective. |
2nd place Tokai Electronics Co., Ltd. Prize money 1 million yen
Application solution: “Monitoring and preventive maintenance by visualizing ‘smell'”
| Comment by Mr. Tsuboi: I appreciated the focus on “smell” and the approach to the real issue of nursing care. I hope that you will continue to work toward practical application. |
1st place Rabbit Co., Ltd. Prize money 2 million yen
Application solution: “‘Gaze simulation AI’ that visualizes the human gaze”
| Commentary by Mr. Tsuboi: We evaluated the uniqueness of the technology that uses the line of sight in such a way, and the unexpectedness of making inferences on the spot with the edge. Also, I think that it is a highly practical technology, so please do your best in the future. |
Ledge.ai Award (Special Award) Salty Star Inc.
Application solution: “Application platform & embedded hybrid database”
| Comments from Mr. Kotaki, President and CEO of Ledge Co., Ltd.: I participated in this review from the perspective of “Is this AI technology accessible to everyone?” I appreciated the high versatility of Saltister’s technology that can be used with various interfaces. I have high hopes for the future. |
In conclusion
At the AI EDGE CONFERENCE & SOLUTION CONTEST, we saw interesting technologies that utilize edge computers. In particular, there were many companies proposing ideas that are conscious of the recent trend of “SDGs” initiatives.
I am very much looking forward to seeing how social issues will be resolved through the evolution of edge computers.
