Background: Advanced manufacturing process involves complex autonomous and robotic systems and monitoring them for possible anomalies are of utmost importance to ensure quality of products as well as safety of the system. While some of the anomalies can be delay-tolerant, the critical anomalies need real-time detection and mitigation, e.g., irregular motion of a robot or sudden presence of an obstacle in the robot’s pathway. Timely detection of these anomalies is essential, and it can be achieved with edge/fog computing which can locally process the data near the sensors that capture the anomaly. Although fixed edge computing is widely used now in the industrial Internet-of-Things (IIoT), this project aims at implementing a drone-edge computing in the advanced manufacturing setup as the drone can move autonomously around the manufacturing systems, also carry sensors, and process the sensor data onboard.

Research Objectives: The objective of this project is to develop a drone-edge computing system that is capable of carrying multiple sensors, e.g., depth sensing and thermal imaging cameras, localization sensors etc., and also can have onboard computing capacity for executing anomaly detection algorithms in real-time while also running its own autonomous navigation and collision avoidance algorithms. In previous work, we developed an obstacle detection with fixed edge computing, and we also developed autonomous drone system for indoor environment as shown in the figure 1. In this project, the drone will have one or more cameras and it needs to fly or hover in the indoor environment autonomously without collision and the drone camera can capture one or more anomalies in the advanced manufacturing system or environment. As this project deals with real-time detection of critical anomalies, the anomaly detection algorithm needs to run on resource-constrained onboard computing along with other algorithms required for the autonomous operation of the drone.

Figure 1: Left: Edge-Computing based obstacle anomaly detection for robotic arm, Right: Modal AI drone capable of AI-based edge computing 

Research Plan: (1) Study the critical anomalies in advanced manufacturing systems, (2) Use a manufacturing system 3D simulation for reproducing anomalies and apply anomaly detection algorithms with 3D videos from the simulation, (3) execute the anomaly detection algorithm onboard along with autonomous drone operations and measure system performance, (4) deploy the drone in a real-world system setup.

REU Student Outcomes: This project fits a student with coding skills (e.g., computer science). REU student will gain hands-on experience in drone technology (hardware and software) as well as learn usage of computer vision and other algorithms. Student will be supervised by Dr. Sabur Baidya (Computer Science and Engineering) and participate in writing conference and journal papers.