Despite the increasing trend for high-fidelity train simulator procurement in the rail industry, current research suggests that simulators are extremely underutilised, which points to ineffective integration arising from one or more disconnects in the management layers. This paper presents a study that set out to: profile the design of driver learning frameworks; investigate how simulators were being integrated; and determine key criteria for simulator acceptance. Data were collected from 61 industry end-users, mostly train drivers, in six rail organisations, and analysed thematically. The findings revealed three Rs that reflected perceptions of poor integration and comprised the dominant end-user evaluation criteria for simulator utility; these were: (i) Reality; (ii) Relevancy; and (iii) Reliability. This paper explores the problem of ineffective integration by using and applying a Robocop allegory, in order to disentangle the dynamic shared between the systemic and cultural influences of the organisation, when new technology is introduced in a highly regulated environment. The paper concludes by presenting three prime directives, triangulated from the study and current literature, to transcend the issues impeding the path of effective simulator integration in rail, and overcome the ‘Robocop problem’ as it applies to this industry.