Deep Learning Assisted Inertial Dead Reckoning and Fusion

TL;DR

This work targets robust navigation for mobile platforms under GNSS outages by exploiting periodic trajectories (PTS) with deep learning. It introduces Mini-QuadNet (MQN), a lightweight network that regresses change in distance and altitude from inertial data, and integrates MQN as external updates into an error-state EKF, enabling neural-inertial fusion both with and without GNSS updates. On real-world quadrotor and mobile-robot datasets totaling 337 minutes, MQN reduces the parameter count by ~88% and improves position RMSE by ~20% versus the baseline QuadNet, while markedly enhancing GNSS-denied dead reckoning (up to ~99% drift reduction) and further boosting accuracy when GNSS is available (up to ~28.8% improvement). The results demonstrate that PTS and neural regression can substantially elevate low-cost inertial navigation performance in both GNSS-enabled and GNSS-denied scenarios, suitable for real-time deployment on limited hardware.

Abstract

The interest in mobile platforms across a variety of applications has increased significantly in recent years. One of the reasons is the ability to achieve accurate navigation by using low-cost sensors. To this end, inertial sensors are fused with global navigation satellite systems (GNSS) signals. GNSS outages during platform operation can result in pure inertial navigation, causing the navigation solution to drift. In such situations, periodic trajectories with dedicated algorithms were suggested to mitigate the drift. With periodic dynamics, inertial deep learning approaches can capture the motion more accurately and provide accurate dead-reckoning for drones and mobile robots. In this paper, we propose approaches to extend deep learning-assisted inertial sensing and fusion capabilities during periodic motion. We begin by demonstrating that fusion between GNSS and inertial sensors in periodic trajectories achieves better accuracy compared to straight-line trajectories. Next, we propose an empowered network architecture to accurately regress the change in distance of the platform. Utilizing this network, we drive a hybrid approach for a neural-inertial fusion filter. Finally, we utilize this approach for situations when GNSS is available and show its benefits. A dataset of 337 minutes of data collected from inertial sensors mounted on a mobile robot and a quadrotor is used to evaluate our approaches.

Figures (8)

• Figure 1: QuadNet architecture as presented in shurin2022quadnet. It consists of convolutional and fully connected layers to output the change in the platform's distance.
• Figure 2: The MQN network architecture. The inertial readings are passed into a series of 1D-CNN and fully connected layers to output the change in distance.
• Figure 3: The neural inertial fusion block diagram. MQN-driven position, as well as GNSS updates, when available, are used as external updates to the navigation filter.
• Figure 4: Examples of PTS trajectories, from our dataset, of a quadrotor (blue) and a mobile robot (orange).
• Figure 5: DJI Phantom 4 RTK equipped with Movella Dot IMUs during our recording campaign.