An Intelligent Software Testing Framework For Cloud-Based Robotic Systems Using Ai And Automation

Abstract

Cloud robotic systems have to contend with so many challenges regarding reliability, scalability, and performance that some conventional software testing methodologies have been assessed as woefully lacking. This paper suggests an AI-based intelligent software-testing framework that allows for automation and effective detection of faults and generation of test cases for modifying coverage in the cloud robotic systems. The proposed intelligent testing framework enjoys the advantages of scalable testing environments via cloud computing, CloudSense Mapping for feature extraction, and then Wavelet Transform, which is the main tool for data preprocessing. Based on sensor data and logs the system categorizes behavior as either normal or anomalous using Long Short-Term Memory (LSTM). The major limitation of the existing manual testing approaches is that it is inefficiently arranged and has a very high operational cost, especially for dynamic cloud environments. This framework addresses all the aforementioned limitations and, during the execution of robotic tasks, exhibits an excellent score, with performance metrics such as accuracy: 98.73%, precision: 97.84%, recall: 98.01%, and F1 score: 97.92%. Actual benefits were 41.6% less cycle time and 38.2% more efficiency as compared to conventional methods of testing. These results indicate that the automated intelligent framework indeed provides significant improvement for the quality assurance processes in cloud-based robotics, towards enabling faster, more reliable, and scalable system validation.