Real-time anthropometric data-driven evaluation method for complex console layout design

A poor complex console layout can easily cause musculoskeletal diseases in operators, which reduces their comfort and efficiency. Existing evaluation methods rely heavily on the subjective experience of experts and do not fully consider the relevant data of operators. This paper explores an anthropometric data-driven evaluation method for designing complex console layouts that incorporates two artificial intelligence algorithms in tandem. Here an Anthropometric Data Extraction (ADE) algorithm was constructed to identify the operator's limb angle and other related data in real time. A multi-stage Artificial Neural Network optimized by Genetic Algorithm (GA-ANN) was designed to accurately and quickly establish the non-mapping relationship between indicators. Taking the manned submersible console operation interface as an example, we evaluated and proved the effectiveness of the proposed method. The ADE algorithm achieved an accuracy of 87.7% in joint point recognition. The GA-ANN algorithm had a better fit and convergence speed. The results indicate that the method could help designers better consider the operator's comfort and perform faster and more accurate design assessments. This could improve the efficiency of design and manufacturing and also provide support for managers in guiding rational posture and task optimization during work.