Harnessing AI worlds for construction skills to close the feedback loop

Vocational training in prisons often fails to equip learners with the skills needed for today’s labour market. Limited access to technology and a shortage of specialist tutors have contributed to poor-quality provision, with many courses rated below standard. Progression from entry-level construction qualifications is low, leaving many workers stuck in semi-skilled roles, despite significant skills shortages in the sector. At the same time, high unemployment and reoffending rates among prison leavers highlight the need for clearer, more effective pathways into work.

This project will develop an AI-enabled training tool for plastering skills, using digital twin technology. Using multi-angle video and sensor-equipped tools, key aspects of a learner’s technique will be captured and analysed. This data will be used to create a digital twin model of each learner’s performance. Tutors will use the system to review performance and provide detailed feedback, while learners will access the tool to practise independently and review their technique.

RMB Consultancy brings together strong sector expertise and partners including CITB, City & Guilds Foundation and the West of England Combined Authority to support delivery and progression. The project will be piloted in prison settings as part of a wider employment pathway, linking learners to real work opportunities. This approach will help learners develop practical plastering skills, support progression into skilled roles, and enable tutors to deliver more consistent, objective feedback. This could improve the quality of vocational training in prisons, strengthen pathways into employment and contribute to reduced reoffending. At scale, the model could be extended to other construction trades and prison-based training programmes.