Leveraging Natural Language Processing for enhanced remote troubleshooting in Product-Service Systems: A case study

In Product-Service System (PSS) offerings is crucial to deliver services in the shortest time to maximize customer satisfaction. One of the first contacts that customers have with the provider is usually through remote assistance delivered via telephone or email. Thus, the development of a structured troubleshooting procedure is fundamental for fast problem identification and resolution. While customers exchange with help desk technicians are saved in a company database, they are often lacking proper structure and, thus, are rarely analyzed. This poses a challenge since aggregated data can provide valuable insights for knowledge extraction and reuse, benefiting PSS lifecycle management and improvement (e.g., enhancing troubleshooting, maintenance service, or PSS design). The paper presents a case study where the textual data collected from the customer ticket database have been analyzed to extract the most recurrent problems and the frequently suggested solutions and improve remote troubleshooting.