IN THE FIELD
TeamSolve’s generative AI-based digital knowledge companion (DKC) enhances Singapore’s water sector
knowledge transfer and access across various organisational levels. The DKC’s core functionalities include knowledge capture and retention, utilising generative AI to mimic and expand human learning processes. It also makes crucial operational information accessible to relevant stakeholders with its knowledge sharing. With DKC’s intelligent standard operating procedure (SOP) support, it assists technicians with necessary information and insights during operations, from troubleshooting to performance comparisons. Over an eight-month period, the DKC was developed, rigorously tested, and iterated upon, with feedback incorporated for continual refinement. Integration into familiar platforms like WhatsApp enhanced user accessibility and experience. It deployed generative AI to preserve and share operator expertise within its teams. It provided with task automation, knowledge preservation, and improved operational efficiency as continuous recording of best practices ensured that operational knowledge is maintained and updated. The implementation of the DKC has led to improvements in streamlined operations. There was a reported reduction in manual tasks like report generation and data entry, leading to more focused and efficient work by technicians. With the AI handling routine information tasks, technicians could concentrate on critical activities, supported by informed, real-time decision-making. Overall, DKC ensured continuous access to institutional knowledge, which is vital for operational continuity, situational responses, and training new staff. It received high approval ratings from technicians, with over 90% reporting greater ease and satisfaction in their role.
in virtual guided troubleshooting for flow monitoring devices. It has quick access to necessary knowledge from documents such as flow meter manuals and reports, and has enabled rapid comparative analysis using knowledge retrieved from technical reports and industry publications to optimise decisions. DKC continuously enriches its knowledge base with tacit knowledge assimilation, retrieving from real-time data and insights from maintenance operations. With the success of the DKC pilot deployment, further recommendations include scaling and integration, along with continued innovation. Enhancements to the DKC architecture and framework as well as investing in advancements like predictive analytics and proactive maintenance insights are positioning it for future expansions in capabilities and applications across other departments and workflows within the organisation.
PUB, Singapore’s national water agency, manages the nation’s water supply with an emphasis on quality, sustainability, and efficiency. PUB has also been exploring innovative methods to preserve and transfer institutional knowledge amid workforce turnover and the retirement of skilled operators. This effort required improved decision-making capabilities within the organisation, especially in the context of workforce turnover and the retirement of skilled operators. PUB carried out a trial using TeamSolve’s generative artificial intelligence (AI)-powered DKC, integrating technologies such as a large language model (LLM) and a knowledge graph (KG) tailored for the maintenance of equipment within the water supply network. This system captures both tacit and explicit knowledge, ensuring continuous Through this pilot deployment, PUB aimed to enhance the preservation and succession of operational knowledge and streamline maintenance processes to maintain the high quality of water services (Image: TeamSolve)
Integration into familiar platforms like WhatsApp enhanced user accessibility and experience (Image: TeamSolve)
The DKC demonstrated its effectiveness in several applications, such as assisting
34 Water & Wastewater Asia | September-October 2024
Made with FlippingBook - professional solution for displaying marketing and sales documents online