Towards a Real-World Robotic Swarm: Consensus Decision-Making, Novel Ground Robots, and Human-Swarm Interaction

Abstract

Towards a real-world robotic swarm, this dissertation focuses on three specific areas within the field of swarm robotics: 1) swarm decision-making; 2) transformable wheel mechanism and its robot embodiment; and 3) adaptive virtual-physical hybrid swarm system and human-swarm user interface (UI). First, this dissertation presents an entropy-based consensus algorithm for a swarm of artificial agents with limited sensing, communication, and processing capabilities. Each agent is modeled as a probabilistic finite state machine a preference for a finite number of options defined as a probability distribution. The most preferred option (called exhibited decision) determines the agent’s state. The state transition is governed by internally updating this preference based on the states of neighboring agents and their entropy-based levels of certainty. By exchanging the exhibited decisions and the certainty values among the locally connected neighbors, swarm agents continuously update their preferences, leading to consensus towards an agreed-upon decision. The presented method is evaluated for its scalability over the swarm size and the number of options and reliability under different conditions. By adopting a classical best-of-N target selection scenario, the algorithm is compared with three existing methods, the majority rule, frequency-based method, and k-unanimity method. The evaluation results show that the entropy-based method is reliable and efficient in both symmetric and asymmetric option consensus problems. Second, this dissertation presents a new passive wheel-leg transformation mechanism based on a unique geared structure, allowing the wheel to transform between two modes, i.e., wheel or leg, potentially adapting to varying ground conditions. It consists of a central gear and legs with partial gears which rotate around the central gear to open or close the legs. When fully closed, the mechanism forms a seamless circular wheel; when opened, it operates in the leg mode. The central gear actuated by the driving motor generates opening and closing motions of the legs without using additional actuator. The number of legs, physical sizes, and the gear ratio between the central gear and the partial gears on the legs are adjustable. This design is mechanically simple, customizable, and easy to fabricate. For physical demonstration and locomotion testing, α-WaLTR, a new adaptive wheel-and-leg transformable robot for versatile multi-terrain mobility is presented. The robot has four passively transformable wheels, each embedded with spring suspension for vibration reduction. These wheels enable the robot to traverse various terrains, obstacles, and stairs while retaining the simplicity in primary control and operation principles of conventional wheeled robots. Unity-based simulations guided the selection of the design variables associated with the transformable wheels. Following the design process, α-WaLTR with an embedded sensing and control system was developed. Experiments showed that the spring suspension on the wheels effectively reduced the vibrations when operated in the legged mode and verified that the robot’s versatile locomotion capabilities were highly consistent with the simulations. Third, this dissertation presents an integrated system for virtual-physical hybrid swarm simulations and human-swarm interaction through mixed reality (MR) UI device. The system has three main components: (1) a virtual module that runs a swarm simulator, (2) a physical module that runs Robot Operating System (ROS), and (3) a UI module that runs human-swarm interface on the UI devices. By establishing the connection of the three components, an interactive hybrid virtual-physical swarm is created to enable collaboration between virtual agents and physical ones. A MR-based human-swarm interface is developed for one or multiple human operators to interact with the hybrid swarm in: defining and sending a task list, offering assistance in target tracking tactics, 1:1 interaction with a physical robot via gesture control and visualizing real-time camera feedback. To prove the flexibility of the proposed UI, two experiments are designed to include different levels of human inputs. Both experiments were conducted with successful results, demonstrating the functionality of the integrated system.