Meidensha: Creating a better future with our people and technology.

scroll
(the Kanogawa Seibu Purification Center)

This content is a reprint of "Meidensha: Creating a better future with our people and technology"
that was posted in the Nikkei Online Edition from July 2018 to August 2018.

(the Kanogawa Seibu Purification Center)

This content is a reprint of "Meidensha: Creating a better future with our people and technology"
that was posted in the Nikkei Online Edition from July 2018 to August 2018.

Meidensha: Creating a better future with our people and technology.

scroll

Stage5 Protecting water infrastructure with AI Looking to the future of sewage treatment plants Making use of on-site data in R&D for failure prediction technologyStage5 Protecting water infrastructure with AI Looking to the future of sewage treatment plants Making use of on-site data in R&D for failure prediction technology

"We are engaged in developing failure prediction technology that will prevent plant emergency stoppages through the continuous collection and analysis of sewage treatment plant data." A sewage treatment plant, the Kanogawa Seibu Purification Center (in Numazu City), is located in the east of Shizuoka Prefecture and is close to Mt. Fuji. On an enormous site of over 18 hectares is a line of buildings for sewage purification. Hidekazu Takeda is visiting the facility for the first time in a while.

At the sewage treatment plant, microbes such as bacteria are used to process sewage to clean water, which is then released to rivers and the sea. Should the central plant have an emergency stoppage, catchment wastewater cannot be properly processed, impacting daily life and prompting concern about resulting environmental pollution.

This sewage treatment plant processes the sewage from a catchment population of around 170,000 people in three cities and two towns including Numazu and Mishima. Approximately 50,000 tons are processed daily. MEIDEN FACILITY SERVICE CORPORATION, a Meiden Group company, is responsible for daily operation and management work. Takeda says, 'We visit the Treatment Plant whenever plant maintenance is required, and install and remove the data collection sensors."

Hidekazu Takeda
Hidekazu TakedaIoT Data Analytics Services Development Section
ICT Product Technology & Service Development Division
ICT Products & Services Management Group
Meidensha Corporation

A blower installed in the Treatment Plant. Making a tremendous noise, the blower sends air into the sewage treatment process, activating the microbes that decompose organic matters.

A number of sensors are mounted on these blowers. "The sensors measure 3-4 kinds of data such as current, temperature and vibration, and send it to the cloud. Based on the collected data, we analyze various phenomena that could lead to failure," Takeda explains.

The blower on which the sensors are installed.

Currently Meidensha is testing remote monitoring and diagnostic services that use Internet of Things (IoT) and Artificial Intelligence (AI). A system is under development that collects data in the cloud and analyzes them with AI to detect any signs of equipment failures or stoppages before they actually happen. Identifying any possible malfunctions or abnormality in advance enables a maintenance technitian to be promptly sent to the site to fix them, reducing plant downtime (stoppages) and greatly improving the accuracy of preventative maintenance..

The data required for system development is continuously collected and accumulated in the cloud.
Meidensha remote monitoring and diagnostic services for sewage treatment plants

Abnormal data collected for a unique AI development

Kosuke Fujita from the IoT Product Development Department, a colleague of Takeda's who joined the company in the same year five years ago, is also involved in development of the remote monitoring and diagnostic services. He is developing an AI-based system that analyzes the electrical current data collected from the sensors and detects abnormality.

Motors for industrial use and for use in water supply and sewage plants, have a long life of 15-20 years. However, motors can deteriorate and fail in a short time for a variety of reasons. Fujita says, "This could happen due to various phenomena such as temperature fluctuation and vibration, but a major cause could also be electrical discharge within the motor due to insulation deterioration." We do conduct regular equipment maintenance to detect these kinds of phenomena, but it is also possible that depending on the timing, not all abnormalites can be detected, leading to malfunction. "So we are developing a system that will detect electrical discharges from the current sensor in real time and diagnose them with AI," says Fujita.

Numeric data for temperature and vibration is relatively easy to analyze. But it is a big job to analyze the discharge data and link it to AI development in order to assess how much impact intermittent electrical discharge has on the motor.

Kosuke Fujita
Kosuke FujitaIoT Data Analytics Services Development Section
ICT Product Technology & Service Development Division
ICT Products & Services Management Group
Meidensha Corporation

The development of AI requires the collection of various types of big data making full use of statistics and machine learning, and the most difficult thing is the collection of abnormal data. Society expects sewage treatment plants to always operate normally. It is not so easy to record actual abnormal values. For that reason, we develop a hypothesis and create pattern data for abnormal status, and then repeat testing. "It is not enough to simply apply data to the hypothesis. Only because Meidensha has been engaged in solving actual customer problems and maintaining the site for many years is it possible to envisage the phenomena that could occur. This is the strength of Meidensha," Fujita tells us. In addition to the technology and know-how accumulated in-house, development speed is increased through open innovation in collaboration with outside universities and other companies.

The biggest issue is the collection of abnormal data.

"Veteran technicians can feel that 'something is wrong' the moment they enter the room where the motor is operating. They can tell from slight differences in sound and vibration if the motor is working properly. However, the number of such veteran technicians is gradually declining. Our customers are hoping that we soon establish this data-based high-precision plant diagnostic technology." Our goal is the development of an AI system that detect signs of failure.

Focused on optimizing regional society with Meidensha ICT

"Japan is facing a globally unprecedented super-aging society, with its total population in 2060 expected to be less than 90 million, of which 40% will be senior citizens." Hirozumi Nishizawa, Planning Department Manager in the ICT Division, emphasizes this sense of crisis. The number of technicians who have supported social infrastructure is declining while the infrastructure built during Japan's high-growth era is rapidly deteriorating. "The efficient operation and maintenance of infrastructure at a time when there are a shortage of technicians and tighter national and municipal budgets is a real issue. We would like to solve these issues with application of ICT," says Nishizawa enthusiastically.

Meidensha believes the company will be able to commercialize its ICT-based remote monitoring and diagnostic services during FY2019. Through the introduction of new services to its municipal customers, the company is aiming to provide an optimal solution that not only brings down costs and reduces manpower but visualizes the amount of chemicals and electricity used in treatment plants.

Hirozumi Nishizawa
Hirozumi NishizawaICT Products & Services Business Planning Division
ICT Products & Services Management Group
Meidensha Corporation

Summary

"We are working to combine ICT technology with the experience and know-how that we have developed at customers sites. It is only heavy electric equipment manufacturers such as Meidensha that can do this in order to maintain social infrastructure by optimizing it and improving its efficiency. Our goal is the optimization of regional society using ICT technology. This means optimization of not only individual facilities such as sewage treatment plants but the optimization of water supply and sewage facilities, power facilities, and transport infrastructure that are vital for people's lives. By connecting them, we are aiming for optimization of the regional society as a whole," says Nishizawa. The social infrastructure spread to all corners of Japan during the high-growth era. The maintenance and management of such infrasturcture has now become a serious problem, as the national and municipal goverments are facing financial difficulties and the number of veteran technicians is declining. There are great expectations for the early introduction of innovative systems that use ICT technology such as AI to control the plant operating budgets and reduce manpower.

Meidensha ICT solutions

Meidensha ICT solutions

Explanation of terminology

Preventative equipment maintenance using IoT
To date, preventative equipment maintenance has primarily been conducted under time-based maintenance, in which inspections, repairs, parts replacement and updates were performed in a certain fixed cycle. However, use of IoT will permit 'condition-based maintenance (CBM)', in which the equipment status is available at any time and parts replacement and repairs, and updates, can be carried out on the basis of predicted wear or failure. If CBM is achieved, the failure rate of the equipment is reduced and maintenance costs less, leading to more efficient operation.
Preventative equipment maintenance using IoTPreventative equipment maintenance using IoT
Collecting and analyzing various data from sensors
*Honorifics omitted.

Meidensha: Creating a better future with our people and technology.

  • Stage1 Powering the excitement.
  • Stage2 Protecting water resources and infrastructure that nurture the community.
  • Stage3 Driving force for a new automotive era.
  • Stage4 Pursuing greater safety and stability in railway transportation.
  • Stage5 AI for water treatment protection.
  • Stage6 Efficient energy management.
  • Stage7 Lightning Protection.
  • Stage8 IoT to protect from heavy rain.
Return to the top of this page