Research
Data-Driven Navigation
ANSFL's team published 26 conference and journal papers in the field of data-driven navigation making our lab a world leader in this topic. The purpose of navigation is to determine the position, velocity, and orientation of manned and autonomous platforms, humans, and animals. Obtaining accurate navigation commonly requires fusion between several sensors, such as inertial sensors and global navigation satellite systems, in a model-based nonlinear estimation framework. Recently, data-driven approaches applied in various fields show state-of-the-art performance, compared to model-based methods. During the last three years, we pioneered, addressed and derived data-driven based navigation algorithms including hybrid learning and end-to-end learning approaches for different platforms and applications. The purpose of those algorithms is to enhance common navigation and estimation tasks and open new possibilities for accurate and robust navigation.
Autonomous Underwater Vehicle Navigation
Fusion between an inertial navigation system (INS) and a Doppler velocity log (DVL) is commonly used in the navigation of autonomous underwater vehicles (AUV). In normal operating scenarios, the navigation accuracy of these systems is sufficient to enable the AUV to reach its goal. However, partial or complete DVL outages may occur if AUVs are operating in complex environments, such as passing over fish and other sea creatures, or where the distance between the DVL and seafloor exceeds the DVL range. In such situations, the navigation solution depends mainly on the INS solution, which drifts over time. In this research, we derived algorithms that facilitate velocity vector estimations for partial and complete DVL outage situations over short time periods using both model and learning approaches. We also demonstrated an approach that improves current velocity estimates in situations of complete beam measurements using DVL measurements and inertial readings (when available).
Information Aided Navigation
The performance of inertial navigation systems is largely dependent on the stable flow of external measurements and information to guarantee continuous filter updates and bind the inertial solution drift. Platforms in different operational environments may be prevented at some point from receiving external measurements, thus exposing their navigation solution to drift. Over the years, a wide variety of works have been proposed to overcome this shortcoming, by exploiting knowledge of the system current conditions and turning it into an applicable source of information to update the navigation filter. Navigating with information is a family of approaches we call information aided navigation. In this research, we broadly classified those approaches into direct, indirect, and model aiding. By matching the appropriate constraint to a given scenario, one will be able to improve the navigation solution accuracy, compensate for the lost information, and uncover certain internal states that would otherwise remain unobservable.