Pumps serve as inconspicuous aides in our daily lives. The traditional maintenance approach for pumps typically involves manual inspections. However, due to insufficient manpower or human-related factors, regular health checkups for pumps are often neglected. Moreover, pumps are frequently submerged in water or liquids with the risk of corrosion, making inspections challenging. The difficulty in monitoring leads to the unfortunate consequence of large-scale disasters when pump abnormalities are eventually identified, significantly impacting the daily lives of the public.
Traditional pumps lack digital connectivity, resulting in reactive emergency repairs when issues arise. The "Pump Smart Cloud System" developed by Syscom shifts from this passive repair model to proactive fault prediction and anomaly alert services. It utilizes sensors such as flow meters and temperature gauges on the pump to collect digital data for the "Smart Panel." The data is then analyzed by the "MiSeeR Edge Fault Prediction and Anomaly Detection System" using AI algorithms to promptly assess the health index of the pump.
Additionally, by employing AI algorithms to predict when pump failures might occur, proactive maintenance measures can be taken to extend the pump's operational lifespan. The "Smart Pump Cloud Service Platform" further enhances user experience by providing advance maintenance notifications and real-time status updates.
■ Intelligent Electrical Panel
The intelligent electrical panel, encompassing circuits and switches, goes beyond basic components, including equipment necessary for edge computing. This apparatus collects crucial data detected by the pump, establishing mechanisms for data collection and analysis. It serves to provide real-time health status updates for the pump and maintenance recommendations.
■ MiSeeR Edge Fault Prediction and Anomaly Detection System
For intelligent pump diagnostics and health assessment, the system relies on a comprehensive evaluation of pump status combined with impeller wear data. The research team utilizes various sensors to extract data features such as flow rate, voltage, current, and temperature using Autoencoder. The Long Short-Term Memory (LSTM) model serves as a key tool to describe flow rate features. By comparing predicted and actual flow rates, the system calculates the pump's health index, establishing an AI computational model for pump health.
■ Intelligent Pump Cloud Service Platform System
The Intelligent Pump Cloud Service Platform offers advance maintenance notices and real-time status detection of pump operation before, during, and after usage. All this real-time data can be proactively communicated to users, enhancing the quality of post-sale services provided by the property owners. In the event of abnormal pump operation, users can utilize the system's maintenance knowledge base (FAQ) to access troubleshooting methods and self-diagnose, reducing the probability of malfunctions and improving user satisfaction.
On June 21, 111, MiTAC joined hands with three other companies, including HOMSEEK Electric, to donate the "Intelligent Pump Cloud Service Platform" to the National Science and Technology Museum (Kaohsiung). Ms. Lu Yuntong, the Director of the Innovative Technology Business Group at MiTAC, represented the company at the donation ceremony and received a certificate of appreciation.
During her speech, Director Lu expressed, "This system solution transcends spatial limitations through the Internet of Things, achieving optimal scheduling and significantly reducing the tremendous losses caused by unexpected downtime. We assist pump manufacturers in upgrading the pump usage experience and support the transformation of the manufacturing industry from being 'manufacturing-oriented' to 'service-oriented'."
The "Intelligent Pump Cloud Service Platform" has successfully alleviated flooding issues for the Kaohsiung Museum of Science and Technology. Leveraging this successful experience, ShenTong Information Co., Ltd.'s AI research and development team actively pursues the development and application of various installations in different domains, including various water treatment projects, smart factories, smart buildings, and more. Mr. Qiu Jinsheng, Vice President of the Innovative Technology Business Group at ShenTong, expressed, "The R&D team has accumulated numerous project experiences in AIoT smart manufacturing solutions. As early as the Smart City Expo in March, the team gained recognition from customers and garnered high interest from visitors. We aim to lead the industry's continuous upgrade and international alignment, assisting customers in creating competitive advantages."