Robot Localization and Navigation through Predictive Processing using LiDAR
This addresses the problem of reliable laser navigation for mobile robots, offering a novel method that eliminates the need for odometry, though it appears incremental as it builds on existing predictive processing concepts.
The paper tackled robot localization and navigation without odometry by using a predictive processing-inspired approach with LiDAR, showing improved state-estimation performance compared to a state-of-the-art particle filter and enabling goal-directed navigation.
Knowing the position of the robot in the world is crucial for navigation. Nowadays, Bayesian filters, such as Kalman and particle-based, are standard approaches in mobile robotics. Recently, end-to-end learning has allowed for scaling-up to high-dimensional inputs and improved generalization. However, there are still limitations to providing reliable laser navigation. Here we show a proof-of-concept of the predictive processing-inspired approach to perception applied for localization and navigation using laser sensors, without the need for odometry. We learn the generative model of the laser through self-supervised learning and perform both online state-estimation and navigation through stochastic gradient descent on the variational free-energy bound. We evaluated the algorithm on a mobile robot (TIAGo Base) with a laser sensor (SICK) in Gazebo. Results showed improved state-estimation performance when comparing to a state-of-the-art particle filter in the absence of odometry. Furthermore, conversely to standard Bayesian estimation approaches our method also enables the robot to navigate when providing the desired goal by inferring the actions that minimize the prediction error.