This thesis explores the design and validation of a human-robot collaboration workstation for a vision-based quality inspection in an industrial-like setting. The goal is to develop a safe, flexible, and efficient system suitable for deployment in environments such as the Smart Factory Lab at Scania CV AB and showing therefore the possibility for an implementation in the real manufacturing. An industrial robot used in a collaborative application equipped with a vision system is integrated into a modular workstation capable of identifying missing components in assembled products. The technical design and vision system of this workstation achieve reliable detection of missing components under industrial conditions and demonstrates adaptability through AI-assisted training and modular software design. Key ergonomic and usability aspects such as station layout, reduced cognitive load, and simplified reprogramming are considered to support flexible task allocation between human and robot. Overall, this workstation is ready to operate in the industrial-like environment and can be implemented in the overall connectivity. Safety is the primary focus of this thesis. The system is designed according to ISO 12100:2010 [1], ISO 10218:2025 [2, 3] and the Machinery Regulation [4] and uses power and force limiting strategies as the only safety feature. Therefore, this thesis guides through the risk assessment methods to minimize hazards. Biomechanical risk validations are performed through both calculation and physical measurement. The results confirmed that under normal operating conditions, all relevant contact forces and pressures remained within the safety thresholds. Proving that only the power and force limiting strategies can be implemented in a human-robot collaboration workstation for a safe vision-based quality inspection. The prototype demonstrates that collaborative vision-based inspection is feasible and safe when operating below head height. The findings support the real-world applicability of the system and provide a strong basis for future work involving full CE compliance, broader defect detection, and further ergonomic studies