Feasibility study of using AI-powered smart insole system to reduce back-pain among older workers in the Norwegian AEC industry

Every year, age-related government expenditure accounts for 20% of the GDP of Europe, and up to 60% of the blue-collar workers aged 50-59 who are put in early retirement say that they would prefer to stay in the workforce. Other than the obvious health hazard, such early retirement puts a burden on the social safety net, retirement benefits, etc. Previous studies have shown that one of the major problems among older workers is back pain, much of which stems from heavy lifting.This study aims to present the benefits and limitations of using a smart insole system. While the technology is not at a mature stage yet, this can be considered as a pilot project to map the expectations of the construction workers, who are the prime beneficiary group. Pressure sensors in the soles generate heatmaps that allow us to identify incorrect posture using an adaptable Artificial Intelligence (AI) lifting engine. This data is then used to evaluate the lifting actions in real-time and preemptively warn the individuals. The target group consists of older blue-collar workers (>60 years) from the Norwegian construction industry, who are engaged in lifting heavy loads manually. From the user’s perspective, registering the lifting action every time can be disruptive for their workflow. From a technical standpoint, the lack of a large enough database for training the model/algorithm limits the accuracy of this AI system. However, as more data is collected from the pilot project, the accuracy as well as reliability will be improved, and future versions can identify such incorrect lifting actions independently without requiring any intervention from the users. Also, we need to be aware that not all types of back pains are immediate, some may show up after weeks or even months of incorrect lifting, and these cannot be registered in the existing system. The lessons learned from this feasibility study will enable older workers to safely continue with their roles while reducing the risk of long-term back injury. Early signs show promise that this novel approach can improve work conditions not just in the construction industry but also extend to other sectors where such manual labor is in use. Keywords: workplace, health and safety, construction industry, artificial intelligence.