Toward the Trustworthiness of Industrial Robotics Using Differential Fuzz Testing





Citation of Original Publication

B. Wang, R. Wang and H. Song, "Toward the Trustworthiness of Industrial Robotics Using Differential Fuzz Testing," in IEEE Transactions on Industrial Informatics, 2022, doi: 10.1109/TII.2022.3211888.


© 2022 IEEE.  Personal use of this material is permitted.  Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.



Intelligent robots are a current application in Industrial Internet of Things (IIoT), with their trustworthiness being a topic of considerable research interest. Vulnerabilities in robot software may affect the trustworthiness of robotics. To detect these vulnerabilities in robot software, this study proposes a differential fuzz testing method. The main idea is to continuously execute test cases for different versions of software packages to detect inconsistencies among outputs and eventually discover vulnerabilities. First, test cases are generated, combining seed generation and mutation, after which the measured model of the packages in RVIZ is built and the generated seeds are executed. The differences among inconsistent outputs are calculated and the causes of the differences analyzed. Two evaluation metrics for the inconsistencies and seeds are presented. This method is applied to the crucial package in ROS-MoveIt!. The results show that the arm.go() of moveit commander has joint angle overflow.