The warehousing sector is among the top when it comes to the risk of developing work-related musculoskeletal disorders (WMSDs), in particular low back pain (LBP). In this sector, LBP is a prevalent issue, due to the nature of the job of lifting and moving (heavy) objects around. The issue has significant implications for the workers’ health, in terms of quality of life. Companies and society feel the consequences in terms of financial costs. This issue could be tackled by introducing smart technology in the form of a smart safety shoe. The concept has been developed by a strategic product design student and the strategic direction has been determined. This project explores the concept further and validates the idea of smart safety shoes to reduce the risk of LBP during manual handling, through technological means.

To understand the problem of LBP in context, extensive literature research was conducted on ergonomics. Understanding what causes it and the current methods to reduce the risks. Further, looking into the possibility of detecting causes through technology. The research results were used to build a prototype for validation of the concept.

The causality of LBP is not easy to point out, as multiple factors (physical, psychosocial, and individual) play a role in its development. Research does conclude that physical factors play a major role, which is related to heavy lifting, repetitiveness, and awkward postures. Manual handling can be performed safely as long as the weight is below 23 kg and correct postures are adopted. Though not all workers adhere to correct posture, and it is hard to track through observational methods.

Postures can be tracked or detected through plantar pressure distribution (PPD), by using pressure sensors. These sensors can be placed within safety shoes and will collect PPD data of workers. The PPD data shows certain patterns and have characteristics that can be linked to different postures. The data can be analysed using machine learning, to automate the process and could be able to give feedback to the user when a risky posture is adopted.

A pressure insole has been prototyped with the conducted research to collect PPD data of different postures (stoop lifting, lifting above shoulder height, and asymmetrical lifting). The collected data were manually analysed to understand how patterns may look like. A machine learning model was made, using a tree algorithm, to analyse the data as well. It can classify all the measured static postures with 100% accuracy. Dynamic lifting data were not analysed by the model yet as it needs additional data preparation. At this point, the concept needs more development to analyse dynamic data and to implement the hardware in the safety shoes.

Based on the results, the core components of the concept have been proven to work and able to detect different postures with great accuracy. The idea of a smart safety shoe that can detect and warn the worker of potential injury is not far-fetched.

This project is the first step in the development of the concept. Due to the complexity of the issue and required knowledge, additional research is needed for the continuation of the project. The posture database has to be set up, improving the machine learning model for dynamic lifting data, hardware design and a live feedback system. With these developments, a smart safety shoe could be brought to market that could improve workers' lives and save additional costs for companies.