Search for: All records

This paper introduces an innovative and streamlined design of a robot, resembling a bicycle, created to effectively inspect a wide range of ferromagnetic structures, even those with intricate shapes. The key highlight of this robot lies in its mechanical simplicity coupled with remarkable agility. The locomotion strategy hinges on the arrangement of two magnetic wheels in a configuration akin to a bicycle, augmented by two independent steering actuators. This configuration grants the robot the exceptional ability to move in multiple directions. Moreover, the robot employs a reciprocating mechanism that allows it to alter its shape, thereby surmounting obstacles effortlessly. An inherent trait of the robot is its innate adaptability to uneven and intricate surfaces on steel structures, facilitated by a dynamic joint. To underscore its practicality, the robot's application is demonstrated through the utilization of an ultrasonic sensor for gauging steel thickness, coupled with a pragmatic deployment mechanism. By integrating a defect detection model based on deep learning, the robot showcases its proficiency in automatically identifying and pinpointing areas of rust on steel surfaces. The paper undertakes a thorough analysis, encompassing robot kinematics, adhesive force, potential sliding and turn‐over scenarios, and motor power requirements. These analyses collectively validate the stability and robustness of the proposed design. Notably, the theoretical calculations established in this study serve as a valuable blueprint for developing future robots tailored for climbing steel structures. To enhance its inspection capabilities, the robot is equipped with a camera that employs deep learning algorithms to detect rust visually. The paper substantiates its claims with empirical evidence, sharing results from extensive experiments and real‐world deployments on diverse steel bridges, situated in both Nevada and Georgia. These tests comprehensively affirm the robot's proficiency in adhering to surfaces, navigating challenging terrains, and executing thorough inspections. A comprehensive visual representation of the robot's trials and field deployments is presented in videos accessible at the following links:https://youtu.be/Qdh1oz_oxiQ andhttps://youtu.be/vFFq79O49dM.

In the field of multi-agent autonomous transportation, such as automated payload delivery or highway on-ramp merging, agents routinely exchange knowledge to optimize their shared objective and adapt to environmental novelties through Cooperative Multi-Agent Reinforcement Learning (CMARL) algorithms. This knowledge exchange between agents allows these systems to operate efficiently and adapt to dynamic environments. However, this cooperative learning process is susceptible to adversarial poisoning attacks, as highlighted by contemporary research. Particularly, the poisoning attacks where malicious agents inject deceptive information camouflaged within the differential noise, a pivotal element for differential privacy (DP)-based CMARL algorithms, pose formidable challenges to identify and overcome. The consequences of not addressing this issue are far-reaching, potentially jeopardizing safety-critical operations and the integrity of data privacy in these applications. Existing research has strived to develop anomaly detection-based defense models to counteract conventional poisoning methods. Nonetheless, the recurring necessity for model offloading and retraining with labeled anomalous data undermines their practicality, considering the inherently dynamic nature of the safety-critical autonomous transportation applications. Further, it is imperative to maintain data privacy, ensure high performance, and adapt to environmental changes. Motivated by these challenges, this paper introduces a novel defense mechanism against stealthy adversarial poisoning attacks in the autonomous transportation domain, termed Reinforcing Autonomous Multi-agent Protection through Adversarial Resistance in Transportation (RAMPART). Leveraging a GAN model at each local node, RAMPART effectively filters out malicious advice in an unsupervised manner, whilst generating synthetic samples for each state-action pair to accommodate environmental uncertainties and eliminate the need for labeled training data. Our extensive experimental analysis, conducted in a Private Payload Delivery Network (PPDN) —a common application in the autonomous multi-agent transportation domain—demonstrates thatRAMPART successfully defends against a DP-exploited poisoning attack with a\(30\% \)attack ratio, achieving an F1 score of 0.852 and accuracy of\(96.3\% \)in heavy-traffic environments.