Biology to Biotechnology: Implementing Characteristics of Bat Navigation on Mobile Robots
Disciplines
Navigation, Guidance, Control, and Dynamics
Abstract (300 words maximum)
Robots are used for a variety of tasks that require a complex series of actions, such as understanding the environment, navigation, and communication, all of which require collaborative and autonomous capabilities. Existing systems accomplish this goal with cameras, radars, and laser scanners, which have two inherent limitations: i they only work in specific lighting/environment conditions, and ii) they generate a massive amount of sensory data, which is often incompatible with platforms with limited onboard computing, memory, and power resources. To address power concerns, we propose the Bio Sonar System (BSS), a unified framework for environmental understanding and communication. The proposed design concept is low-cost, quick, and appropriate for real-time applications with changing lighting and environments. The proposed method is based on how bats navigate in a swarm using lightweight transducers such as a nose (or mouth) and two ears. In a complex environment like caves, such transducers are sufficient for achieving excellent inter and intra communication. As a result, the proposed BSS method mimics bats' ability to sense geometry of complex unstructured natural environments by utilizing distributed sensing and following adaptively nominated swarm leaders. To navigate and complete the task, BSS allows robots to detect the speed and distance of nearby objects, as well as their 3D shape, in order to effectively echo-locate the environment. 1) Data Acquisition is one of the project's specific goals. ii) Scene Recognition and Navigation We want to create a lightweight 3D Convolutional NeuralNetwork (3DCNN) for real-time detection and localization of nearby objects. Each robot with a pre-trained 3DCNN model would use cutting-edge navigation algorithms for collaborative sensing, route planning, and tracking in previously unseen real-world scenarios. iii) The ability to communicate. BSS sends out sound waves and uses echo signals to understand the environment. Allowing for effective mobile robot navigation in unmapped environments.
Academic department under which the project should be listed
SPCEET - Robotics and Mechatronics Engineering
Primary Investigator (PI) Name
Muhammad Hassan Tanveer
Biology to Biotechnology: Implementing Characteristics of Bat Navigation on Mobile Robots
Robots are used for a variety of tasks that require a complex series of actions, such as understanding the environment, navigation, and communication, all of which require collaborative and autonomous capabilities. Existing systems accomplish this goal with cameras, radars, and laser scanners, which have two inherent limitations: i they only work in specific lighting/environment conditions, and ii) they generate a massive amount of sensory data, which is often incompatible with platforms with limited onboard computing, memory, and power resources. To address power concerns, we propose the Bio Sonar System (BSS), a unified framework for environmental understanding and communication. The proposed design concept is low-cost, quick, and appropriate for real-time applications with changing lighting and environments. The proposed method is based on how bats navigate in a swarm using lightweight transducers such as a nose (or mouth) and two ears. In a complex environment like caves, such transducers are sufficient for achieving excellent inter and intra communication. As a result, the proposed BSS method mimics bats' ability to sense geometry of complex unstructured natural environments by utilizing distributed sensing and following adaptively nominated swarm leaders. To navigate and complete the task, BSS allows robots to detect the speed and distance of nearby objects, as well as their 3D shape, in order to effectively echo-locate the environment. 1) Data Acquisition is one of the project's specific goals. ii) Scene Recognition and Navigation We want to create a lightweight 3D Convolutional NeuralNetwork (3DCNN) for real-time detection and localization of nearby objects. Each robot with a pre-trained 3DCNN model would use cutting-edge navigation algorithms for collaborative sensing, route planning, and tracking in previously unseen real-world scenarios. iii) The ability to communicate. BSS sends out sound waves and uses echo signals to understand the environment. Allowing for effective mobile robot navigation in unmapped environments.