A series of research studies have been conducted over the course of five years venturing into the fields of in-service inspections (ISI) in nuclear power plants (NPPs) and inspection of manufactured components to be used for permanent nuclear waste disposal. This paper will provide an overview of four research studies, present selected experimental results and suggest ways for optimization of the NDT process, procedures, and training. The experimental results have shown that time pressure and mental workload negatively influence the quality of the manual inspection performance. Noticeable were influences of the organization of the working schedule, communication, procedures, supervision, and demonstration task. Customized Failure Mode and Effects Analysis (FMEA) was used to identify potential human risks, arising during acquisition and evaluation of NDT data. Several preventive measures were suggested and furthermore discussed, with respect to problems that could arise from their application. Experimental results show that implementing human redundancy in critical tasks, such as defect identification, as well as using an automated aid (software) to help operators in decision making about the existence and size of defects, could lead to other kinds of problems, namely social loafing and automation bias that might affect the reliability of NDT in an undesired manner. Shifting focus from the operator, as the main source of errors, to the organization, as the underlying source, is a recommended approach to ensure safety.