In the manufacturing industry, maintenance after a sudden machine stop is of utter importance, as the standstill of a machine reduces the profit of the manufacturing company. These sudden stoppages may not be preventable by Predictive or Proactive Maintenance. Therefore, approaches to Reactive Maintenance need to be improved. This paper aims to solve this problem by proposing a Maintenance Assistance System that uses Root Cause Analysis (RCA) to identify the root causes of downtime and provide assistance in solving the underlying problem. Moreover, through a human-in-the-loop approach, one of the weaknesses of conventional RCA, namely the lack of data for decision making, is improved. In addition, two different algorithms for decision making were developed and evaluated with a focus on accuracy of recommendation and efficiency for the operator. The results provide an indication that this type of maintenance assistance has a positive effect on machine downtime and that the decision-making algorithms have certain trade-offs between accuracy and efficiency.