“Smart Echo”, a device using AI and ultrasonic waves to distinguish between male and female fishes, and “TWINDS-T”, a wireless temperature sensor, developed with advanced technical support from Tohoku UniversityTohto C-tech Corp
“Smart Echo” BX series. Male and female can be distinguished instantly by pointing the waves from the device at the abdomen of a fish
Points to be noted
- Promote development by understanding industry needs with advanced technology support from Tohoku University
- Capture ultrasound images and improve business efficiency through AI learning with “Smart Echo”
- Contribute to the reduction of defective products by visualizing the temperature inside molds with “TWINDS-T”
Founded in 2002, Tohto C-tech Corporation was accepting orders to develop embedded software for products. After the Lehman shock, they focused on developing in-house products with advanced technical support from Tohoku University in addition to the know-how they had collected.
“Smart Echo” was one such accomplishment.
The impetus for development goes back to the 2011 Great East Japan Earthquake. Specially-appointed professor Mitsuru Kano of Intelligent Information System Research Center (IIS Research Center), Tohoku University conducted interview surveys around the Sanriku coastal area, which was damaged by the tsunami, and identified a number of research themes for contribution to reconstruction, and the company started developing a device to distinguish whether “Harako” of Pacific cod is milt or fish eggs from these themes.
Even veteran fishermen find it difficult to distinguish between male and female fishes based on appearance and the few “Experts” at present have to identify one by one based on their experience.
The company started the research and development of the device to distinguish between male and female fishes from 2012. They first attempted with a method using the spectral characteristics of X-rays and infrared rays which was not successful, and from 2014, they switched over to the development using ultrasonic waves, which is the research seed of Professor Yoshifumi Saijo of the Graduate School of Biomedical Engineering, Tohoku University.
Smart Echo can capture images of the abdomen through ultrasonic waves by pointing the device at the abdomen of a fish. AI was made to learn the images of the abdomen, and male and female fishes could be distinguished. In February 2019, they started shipping the samples to fishery cooperatives, and fish processing companies, etc. The company is preparing for a full-scale launch.
At the same time, they are also developing a device to distinguish between male and female fishes for Salmon and are also promoting the development of the method for different fish species.
There are two models of Smart Echo.
In BX series, male and female fishes can be distinguished with just this device. It is completely waterproof and can be used on the ship and fishing grounds where there are chances of water splashing.
The other type is the SX series using which the images of the fish abdomen can be viewed directly with a smartphone. The device can also be used for research in aquaculture facilities and fishery test stations, and other uses of the invention apart from distinguishing male and female fishes can also be expected.
The fishery industry is also facing severe labor shortage like many primary industries, and there is a need to manage the fishery industry with few people by using IT, etc. If Smart Echo is used extensively, the work to distinguish between male and female fishes will be simplified improving work efficiency.
The company also has another product called “TWINDS-T”.
A maximum of 8 thermocouples is arranged inside the molds of die casting machines and temperature is measured every 0.1 seconds with TWINDS-T. The measurement results can be checked on a computer from work management rooms, which are at remote locations, using Wi-Fi. A graph is displayed on the computer to show the changes in temperature and can be monitored in real time.
Temperature control inside molds is known to be very important in metal casting such as die casting. Up to now, there was no method to measure the temperature that reaches several hundred degrees inside the molds, and the process relied on the experience of skilled technicians.
The machine intelligence study group “MITOOS (Machine Intelligence Tohoku Society)” works on identifying and solving the problems faced by the industry with industry-academia collaboration between Tohoku University and 40 to 50 small and medium-sized enterprises in the areas around Sendai City. Measurement of the temperature inside molds was also put up as a requirement to MITOOS, and the company and Takumi Solutions Co., Ltd., which are member companies, cooperated with the development.
TWINDS-T which was developed can establish a temperature index when non-defective products are manufactured by collecting data. If the process is monitored regularly, an alarm notifies abnormalities when there is a deviation from the index, and the number of defective products can be reduced by stopping and adjusting the machines so that the defective products do not move to the next process.
The company intends to increase the number of temperature measurement points from 8 to 16 and 32 points. They are planning to expand the range of application to not only die casting but also other industries by selecting the most suitable thermocouple according to the object to be measured.
The company aims contribute to the development of all industries with advanced technology including AI.
Interview Date: February 22, 2019
We offer the best solution or product to solve the problem each customer has by applying AI and image processing technologies provided in academic-industrial cooperation with Tohoku University. We are also running a solution business using experience and know-how of leading edge software development in fields for in-vehicle equipment and semiconductor manufacturing equipment.