The work aims to explore the underlying meaning behind each behavior during the use of intelligent mowing robots, reconstruct the operational process, product functions, and component systems, and improve product usability and user experience. Non-participatory observation and in-depth interviews were conducted to analyze users' behaviors when using mowing robots. The SAPAD framework was introduced to uncover the mapping relationship between "behavior-object-meaning", and the general meanings were obtained by clustering. The AHP method was used to calculate the weights of general meaning clusters, and core meaning clusters at multiple levels were obtained through screening and clustering to identify users' core needs. By analyzing the mapping between "multi-level core meaning cluster-product", the component system was reconstructed. The design of an intelligent mowing robot based on the SAPAD-AHP method can more accurately capture the core needs of users, make the functional design more reasonable, and help designers to create more competitive solutions. It provides a reference and guidance for the design and development of similar intelligent robot products.