Accuracy and efficiency are extremely critical factors for large real-time control applications. A small oversight can cause catastrophic failure of a real-time system. Thus, these applications have to be tested meticulously to prevent any catastrophe that might occur. But, testing these applications exhaustively is not tractable, mainly due to the inherent complexity of the applications and also the huge amount of inputs and outputs that these applications involve. In order to save valuable amounts of time and resources, automated testing is imperative. Also, quantitative metrics have to be provided that assess the existing quality of the system and help increase the confidence in the user towards the software. However, to improve the overall quality of the software, additional focused testing needs to be done. The work in this thesis involves providing specific test suggestions that help the user conduct thorough and precise domain tests based on the knowledge of the various parameters used in previous test runs. The information about the defective portions of the input domain is provided by dividing the input range into percentiles, which is referred to here as bucketing. The goal is to expose the exact inputs causing the defects and the range of inputs that have been lightly tested or left untested during previous tests. A Reliability Analysis Test Tool (RATT) was developed to implement these test suggestions.