The pursuit of reliable and high-precision indoor positioning has become increasingly critical with the widespread deployment of Unmanned Autonomous Systems (UAS) across smart cities. While Wi-Fi Round-Trip-Time (RTT) technology offers promising absolute positioning capabilities, it faces challenges from signal interference and processing delays. Similarly, LiDAR-inertial odometry (LIO) systems provide accurate relative positioning, but suffer from cumulative drift over time. Although existing methods have explored loosely coupled technologies, they process sensor data separately, failing to fully exploit the complementary strengths of different sensors. This research pioneered a tightly-coupled RTT/LIO framework, encompassing novel factor graph formulations that ensure consistency between RTT and LiDAR observations, alongside LiDAR-aided RTT outlier detection and exclusion. Furthermore, we developed an innovative approach to estimate the positions of unknown access points (AP) by using prior trajectory and RTT observations. AP position estimation is based on kernel density estimation (KDE) and geometric diversity constraints (GDC) with the help of an adaptive RANSAC-based fault detection algorithm. Compared to RTT-only implementations, state-of-the-art LIO systems, and conventional loosely coupled approaches, our method demonstrated error reductions of 20-80% in extensive experiments. The implementation of our proposed methodology has been made publicly available on GitHub. The video Bilibili is also shared to display our research.

Reinforcement learning (RL) has shown great potential for autonomous navigation within internet of things (IoT) environments, where various and changing uncertainties pose significant challenges for safe, real-world deployment. Existing safe RL methods typically employ heuristic constraints while neglecting the combined impact of multiple uncertainty sources, reducing robustness and interpretability. Drawing on concepts from global navigation satellite system (GNSS) integrity monitoring, this paper proposes an epistemic integrity risk monitoring reinforcement learning (EIRM-RL) framework to enable trustworthy autonomous navigation under uncertainty. EIRM-RL extends the GNSS protection level concept to RL by utilizing an assembled world model that quantifies and incorporates sensor noise, systematic bias, and epistemic uncertainty. Furthermore, the framework continuously monitors a dynamic epistemic risk probability, which is incorporated into policy optimization as an adaptive safety constraint via Lagrangian duality. This method enables the agent to proactively avoid hazards and effectively balance safety and performance, even in highly uncertain environments. Extensive experiments demonstrate that EIRM-RL achieves superior success rates, collision avoidance, and robustness compared to state-of-the-art safe RL methods, while maintaining high efficiency.

Wearable devices, such as smartwatches, have been widely used by the public for health monitoring, exercise tracking, and message alerts. Regarding localization applications, although smartwatches have been explored by many researchers, current solutions typically require combining data from multiple sensors or multiple mobile devices for reliable position estimation, which is inconvenient for practical usage. In this paper, we propose a novel indoor localization method that relies solely on an inertial measurement unit (IMU) of a smartwatch. First, we design a smartwatch-based neural inertial odometry (WNIO) that adapts to multiple motion patterns and achieves low-drift dead reckoning. Second, we detect interactions between the arm and the environment, such as opening a door, to formulate absolute position measurements that help constrain error drift. A multi-hypothesis Kalman filter (MHKF) is employed to ensure reliable matching and localization. We conducted real-world experiments to validate the effectiveness of the proposed method. The results demonstrate that our method achieves reliable dead reckoning across various motion patterns, including walking and running at different speeds. Furthermore, the method effectively mitigates error drift without relying on any external signals. The video (BiliBili) is also shared to display our research.

Global navigation satellite system (GNSS) positioning can be significantly degraded due to multipath and non-line-of-sight (NLOS) signals in urban areas. Cellular vehicle-to-everything (C-V2X) technology provides new opportunities to enhance GNSS performance from a single intelligent vehicle by leveraging roadside GNSS (RSG) and C-V2X. Inspired by this, we propose an RSG-aided GNSS/LiDAR/IMU (RSG-GLIO) method to achieve reliable odometry and mapping, which leverages the high-quality double-differenced (DD) measurements provided by nearby RSG, effectively mitigating shared random errors such as multipath and NLOS. Our RSG-GLIO first estimates the absolute state of the vehicle using onboard sensors. Utilizing this initial positioning estimate, the proposed method introduces a coarse-to-fine selection scheme to identify consistent DD observations from available RSG measurements. Finally, the consistent roadside DD constraints are jointly optimized into factor graph optimization (FGO). Static and dynamic data are extensively evaluated using multiple RSG receivers deployed in the Hong Kong C-V2X testbed to evaluate the effectiveness of roadside-aided positioning. The results demonstrate a significant 36.6% improvement in terms of absolute positioning accuracy compared to the state-of-the-art GLIO method. Furthermore, we showcase the potential for employing RSG as low-cost base stations in dense urban areas. The data of our work is publicly accessible at https://github.com/DarrenWong/RSG-GLIO.

The use of mobile devices for indoor localization has proven to be a convenient solution for pedestrians in Internet of Things (IoTs) applications. Radiofrequency (RF) signals, including Wi-Fi, Bluetooth, and others, are among the most commonly used sources. However, their availability cannot be guaranteed in all scenarios. Although pedestrian dead reckoning (PDR) using an inertial measurement unit (IMU) provides a self-contained positioning solution, it is susceptible to error accumulation due to heading uncertainties and varying motions. This article presents a low-drift 3-D deep-inertial odometry (DIO) method for indoor pedestrian localization using an IMU. The proposed approach employs a neural network to regress speeds within the human body frame, ensuring that the speeds are unaffected by absolute heading. These regressed speeds are integrated with inertial navigation to determine position. To enhance accuracy, the method incorporates an invariant extended Kalman filter (InEKF)-based integration for state estimation. Additionally, a learned height is included in the filter to improve 3-D position estimation. The performance of the proposed method is validated through real-world tests in various environments. Results demonstrate that the proposed method outperforms traditional PDR, robust neural inertial navigation (RONIN), and EKF-based techniques. Furthermore, this article examines the method from multiple perspectives, highlighting its strengths in addressing heading drift and varying motions, as well as the impact of height constraints and behavior-based position corrections. The video (YouTube) or (BiliBili) is shared to showcase our work.

In the field of autonomous driving, the micro-electromechanical systems (MEMS)-based vehicle navigation usually adopts multi-sensor integrated navigation to achieve high-precision positioning. However, due to the complex environments, the accuracy and reliability of navigation sensors may be significantly reduced. To address these challenges, the interacting multiple model (IMM)-based strapdown inertial navigation system/global navigation satellite system/odometer/non-holonomic constrain (SINS/GNSS/OD/NHC) integrated navigation is adopted. Unfortunately, due to the use of traditional state-space model (SSM), the existing IMM estimation methods often suffer from poor estimation consistency, and the mounting error angle will also make OD/NHC subfilter models affect estimation consistency during the interaction process. Moreover, complex environments lead to frequent model switching, and relying on inaccurate model probabilities may cause significant fluctuations in the subfilter outputs, thereby reducing estimation accuracy. In contrast, the proposed IMM estimation method constructs Lie group-based subfilter SSM, which improves estimation consistency. Additionally, the velocity-bias-based mounting error angle estimation method is proposed by using variational Bayesian (VB) techniques, which further refines the OD/NHC models in IMM. On the other hand, a dynamic likelihood adaptive mechanism (DLAM) is introduced to improve reliability and mitigate the negative effects of frequent switching. Simulation and field test results demonstrate that the proposed velocity-bias-based mounting error angle estimation method has fast convergence speed and high convergence accuracy. Additionally, the proposed Lie group-based reliable IMM estimation method has better accuracy and robustness in complex environments as compared to the existing IMM estimation methods.

Global Navigation Satellite Systems (GNSS) are crucial for intelligent transportation systems (ITS), providing essential positioning capabilities globally. However, in urban canyons, the GNSS performance could significantly degraded due to the blockage of direct GNSS signals. The pseudorange measurements are largely affected and the conventional model of weighting observations is not suitable in urban canyons. This paper addresses these challenges by integrating Artificial Intelligence (AI), specifically deep learning, into GNSS positioning process to enhance positioning accuracy. Traditional methods have primarily focused on pseudorange correction due to the absence of ground truth for weight estimation. In response, we propose an innovative indirect training approach using deep learning to optimize both pseudorange bias and weight estimation, aiming to minimize the positioning errors. To support this integration, we developedpyrtklib, a Python binding for the open-source RTKLIB tool, bridging the gap between traditional GNSS algorithms, typically developed in Fortran or C, and modern Python-based AI frameworks. Comparative analyses demonstrate that our method surpasses established tools like goGPS and RTKLIB in positioning accuracy, marking a significant advancement in the field. The source code of tightly coupled deep learning and GNSS integration, along with pyrtklib, is available on GitHub at https://github.com/ebhrz/TDL-GNSS and https://github.com/IPNL-POLYU/pyrtklib.

Global navigation satellite system (GNSS) positioning in urban environments suffers from significant accuracy degradation due to non-line-of-sight (NLOS) signal receptions. Existing correction methods, such as 3D model-aided and 3D LiDAR-aided GNSS, lack signal direction information and typically construct candidate reflection paths by exhaustively searching over possible reflection surfaces or azimuth angles, and selecting the final path based on the shortest-path assumption. However, this assumption is often invalid in dense urban canyons. To address this limitation, we propose a novel GNSS NLOS correction method that uses Doppler shift measurements to infer signal directional information, which is integrated with real-time point cloud mapping to reconstruct the actual signal reflection path actively. This approach allows us to directly track signal reflection, eliminating the need for exhaustive candidate generation and the shortest-path assumption. Experiments conducted on datasets collected in urban canyons demonstrate the effectiveness of the proposed method. Results show that the method achieves over 90% correction availability for NLOS signals, leading to more than 50% improvement in 3D GNSS positioning accuracy.

In the low-altitude economy, ensuring the safe and agile flight of unmanned aerial vehicles (UAVs) in dynamic obstacle environments is essential for expanding interactive applications like parcel delivery. While deep reinforcement learning (DRL) shows promise for UAV motion planning and control, its trial-and-error exploration often struggles to ensure both agility and safety, especially under uncertain observational noise. Therefore, this paper proposes a deep confidence-enhanced reachability policy optimization (DCRPO) framework. By integrating safe DRL with nonlinear model predictive control (NMPC), DCRPO achieves high-level safety decisions, complex real-time joint planning and control for UAVs. Furthermore, we develop a deep confidence-enhanced reachability guarantee that constructs a set of stochastically forward-reachable planned trajectories under uncertainty, enabling robust safety collision probability certifications. This safe reachability mechanism adaptively selects belief space actions from planned actions to interact with the environment, further enhancing safety and reducing training time. In extensive experiments of UAVs traversing a fast-moving rectangular gate, the proposed method outperforms other state-of-the-art baseline methods under varying environments in terms of operational robustness. Furthermore, the proposed method significantly reduces overall collision violations and training time, greatly improving both training safety and efficiency. The demonstration video (https://youtu.be/7xkp9U7FSJg) and the source code (https://github.com/ZyyFLY/DCRPO) are also provided.

Accurate and safety-quantifiable localization is of great significance for safety-critical autonomous systems, such as Autonomous ground vehicles (AGVs) and autonomous aerial vehicles (AAVs). The visual odometry-based method can provide accurate positioning in a short period but is subject to drift over time. Moreover, the quantification of the safety of the localization solution (the error is bounded by a certain value) is still a challenge. To fill the gaps, this paper proposes a safety-quantifiable line feature-based visual localization method with a prior map. The visual-inertial odometry provides a high-frequency local pose estimation, which serves as the initial guess for the visual localization. By obtaining a visual line feature pair association, a foot point-based constraint is proposed to construct the cost function between the 2D lines extracted from the real-time image and the 3D lines extracted from the high-precision prior 3D point cloud map. Moreover, a global navigation satellite system (GNSS) receiver autonomous integrity monitoring (RAIM) inspired method is employed to quantify the safety of the derived localization solution. Among that, an outlier rejection (also well-known as fault detection and exclusion) strategy is employed via the weighted sum of squares residual with a Chi-squared probability distribution. A protection level (PL) scheme considering multiple outliers is derived and utilized to quantify the potential error bound of the localization solution in both position and rotation domains. The effectiveness of the proposed safety-quantifiable localization system is verified using the datasets collected by AAV and AGV in indoor and outdoor environments, respectively. The open-source code is available at https://github.com/ZHENGXi-git/SafetyQuantifiable-PLVINS

Pedestrian location tracking in emergency responses and environmental surveys of indoor scenarios tend to rely only on their own mobile devices, reducing the usage of external services. Low-cost and small-sized inertial measurement units (IMU) have been widely distributed in mobile devices. However, they suffer from high-level noises, leading to drift in position estimation over time. In this work, we present a graph-based indoor 3D pedestrian location tracking with inertial-only perception. The proposed method uses onboard inertial sensors in mobile devices alone for pedestrian state estimation in a simultaneous localization and mapping (SLAM) mode. It starts with a deep vertical odometry-aided 3D pedestrian dead reckoning (PDR) to predict the position in 3D space. Environment-induced behaviors, such as corner-turning and stair-taking, are regarded as landmarks. Multi-hypothesis loop closures are formed using statistical methods to handle ambiguous data association. A factor graph optimization fuses 3D PDR and behavior loop closures for state estimation. Experiments in different scenarios are performed using a smartphone to evaluate the performance of the proposed method, which can achieve better location tracking than current learning-based and filtering-based methods. Moreover, the proposed method is also discussed in different aspects, including the accuracy of offline optimization and proposed height regression, and the reliability of the multi-hypothesis behavior loop closures. The video (YouTube) or (BiliBili) is also shared to display our research.

LiDAR-based localization is valuable for applications like mining surveys and underground facility maintenance. However, existing methods can struggle when dealing with uninformative geometric structures in challenging scenarios. This paper presents RELEAD, a LiDAR-centric solution designed to address scan-matching degradation. Our method enables degeneracy-free point cloud registration by solving constrained ESIKF updates in the front end and incorporates multisensor constraints, even when dealing with outlier measurements, through graph optimization based on Graduated Non-Convexity (GNC). Additionally, we propose a robust Incremental Fixed Lag Smoother (rIFL) for efficient GNC-based optimization. RELEAD has undergone extensive evaluation in degenerate scenarios and has outperformed existing state-of-the-art LiDAR-Inertial odometry and LiDAR-Visual-Inertial odometry methods.

Collaborative state estimation using different heterogeneous sensors is a fundamental prerequisite for robotic swarms operating in GPS-denied environments, posing a significant research challenge. In this paper, we introduce a centralized system to facilitate collaborative LiDAR-ranging-inertial state estimation, enabling robotic swarms to operate without the need for anchor deployment. The system efficiently distributes computationally intensive tasks to a central server, thereby reducing the computational burden on individual robots for local odometry calculations. The server back-end establishes a global reference by leveraging shared data and refining joint pose graph optimization through place recognition, global optimization techniques, and removal of outlier data to ensure precise and robust collaborative state estimation. Extensive evaluations of our system, utilizing both publicly available datasets and our custom datasets, demonstrate significant enhancements in the accuracy of collaborative SLAM estimates. Moreover, our system exhibits remarkable proficiency in large-scale missions, seamlessly enabling ten robots to collaborate effectively in performing SLAM tasks. In order to contribute to the research community, we will make our code open-source and accessible at https://github.com/PengYu-team/Co-LRIO.

In this paper, a new UAV dataset is presented to support UAV research, such as high-precision positioning and dynamic calibration. The presented dataset is divided into two categories based on different research needs. The first category of the dataset contains visual, inertial, and motor encoder information collected in the indoor motion capture room. This dataset provides accurate ground truth generated by motion capture, which is suitable for the study of UAV dynamics model. The other category of the dataset is collected in a variety of complex outdoor scenarios, and the multi-sensor fusion localization algorithm is used to generate high-precision ground truth trajectory, this category of the dataset could be used for research in UAV positioning and scene reconstruction in complex environments. In short, a total of nine sequences of the dataset are provided. More importantly, the timestamps of raw measurements in each sequence are well synchronized and accurately calibrated. The dataset also provides accurate extrinsic and intrinsic parameters and ground truth trajectories.

Fingerprint localization is a critical method for indoor positioning that has garnered considerable attention. Traditional methods for constructing fingerprint databases and matching algorithms frequently exhibit inefficiencies and limitations, which can undermine both the accuracy and the robustness of localization systems. This paper introduces an innovative indoor pervasive localization method leveraging deep learning. We employ a hybrid system of foot-mounted positioning devices and smartphones to efficiently create a universal fingerprint database, subsequently utilizing LSTM-based deep learning methods for accurate pedestrian location matching. Experiments conducted within a standard academic building demonstrate that our proposed method can more accurately map the indoor movement trajectories of pedestrians. The localization results indicate a horizontal accuracy of 2.5 meters and a vertical accuracy of 0.2 meters. Notably, our method shows a 10% improvement in horizontal accuracy and an 18% improvement in vertical accuracy over WiFi-only approaches. Furthermore, when compared to the classical Random Forest models, our method achieves performance enhancements of 20% in horizontal and 15% in vertical accuracy.

Smooth and accurate global navigation satellite system (GNSS) positioning for pedestrians in urban canyons is still a challenge due to the multipath effects and the non-line-of-sight (NLOS) receptions caused by the reflections from surrounding buildings. Factor graph optimization (FGO) attracts more and more attention in GNSS society for improving urban GNSS positioning by effectively exploiting the measurement redundancy from historical information to resist the outlier measurements. Unfortunately, the FGO-based GNSS standalone positioning is still challenged in highly urbanized areas. As an extension of the previous FGO-based GNSS positioning method, the potential of the pedestrian dead reckoning (PDR) model in FGO to improve the GNSS standalone positioning performance in urban canyons is exploited in this paper. Specifically, the relative motion of the pedestrian is estimated based on the raw acceleration measurements from the onboard smartphone inertial measurement unit (IMU) via the PDR algorithm. Then the raw GNSS pseudorange, Doppler measurements, and relative motion from PDR are integrated using the FGO. Given the context of pedestrian navigation with a small acceleration most of the time, a novel soft motion model is proposed to smooth the states involved in the factor graph model. This paper verified the effectiveness of employing the PDR model in FGO step-by-step through two datasets collected in dense urban canyons of Hong Kong using smartphone-level GNSS receivers. The comparison between the conventional extended Kalman filter, several existing methods, and FGO-based integration is presented. The proposed method shows better results than the conventional FGO method in all test datasets, with at least a 22% decrease in the mean value of positioning error. The proposed method reduces the average localization error from 31.64 m to 18.51 m in a deep urban area.

Indoor localization by smartphones has indicated its promising application prospect in daily life. Smartphone-based pedestrian dead reckoning (PDR) is a common method to obtain the locations. However, PDR suffers from position error accumulation. Although radio frequency (RF) and indoor map can be utilized to restrain the error drift, it requires the prior deployment of facilities or information, which is unsuitable for unknown environments. This paper proposes a factor graph optimization (FGO)-based indoor pedestrian simultaneous localization and mapping (SLAM) with probabilistic exact activity loop closures using a smartphone. In this paper, the smartphone built-in inertial measurement unit (IMU) is solely used to achieve SLAM, in which the human turning activity is regarded as the landmark. Repeatedly observed activities are then used to form loop closures to restrain the drift. FGO is first utilized to formulate pedestrian IMU-only SLAM, which achieves better estimation accuracy than the filter-based method. Moreover, multi-hypothesis tracking is employed to deal with ambiguous data association. During the turning, key points are defined and mutually matched to form exact loop closures to improve estimation accuracy. Simulations and experimental tests are both done to evaluate the performance of the proposed method.

This paper proposes a 3D LiDAR aided global navigation satellite system (GNSS) non-line-of-sight (NLOS) mitigation method due to both static buildings and dynamic objects. A sliding window map describing the environment of the ego-vehicle is first generated, based on real-time 3D point clouds from a 3D LiDAR sensor. Subsequently, the NLOS receptions are detected based on the sliding window map using a proposed quick searching method which eliminates the reliance on the initial guessing of the position of the GNSS receiver. Instead of directly excluding the detected NLOS satellites from further estimating the position, this paper rectifies the pseudo-range measurement model by (1) correcting the pseudo-range measurements if the reflecting point of the NLOS signals is detected within the sliding window map, and (2) remodeling the uncertainty in the NLOS pseudo-range measurement using a novel weighting scheme. The performance of the proposed model was experimentally evaluated in several typical urban canyons in Hong Kong using an automobile-level GNSS receiver. Furthermore, the potential of the proposed NLOS mitigation method in GNSS and the integration of inertial navigation systems were evaluated via factor graph optimization.

Global navigation satellite systems (GNSS) are one of the utterly popular sources for providing globally referenced positioning for autonomous systems. However, the performance of the GNSS positioning is significantly challenged in urban canyons, due to the signal reflection and blockage from buildings. Given the fact that the GNSS measurements are highly environmentally dependent and time-correlated, the conventional filtering-based method for GNSS positioning cannot simultaneously explore the time-correlation among historical measurements. As a result, the filtering-based estimator is sensitive to unexpected outlier measurements. In this paper, we present a factor graph-based formulation for GNSS positioning and real-time kinematic (RTK). The formulated factor graph framework effectively explores the time-correlation of pseudorange, carrier-phase, and doppler measurements, and leads to the non-minimal state estimation of the GNSS receiver. The feasibility of the proposed method is evaluated using datasets collected in challenging urban canyons of Hong Kong and significantly improved positioning accuracy is obtained, compared with the filtering-based estimator.

Absolute positioning is an essential factor for the arrival of autonomous driving. At present, GNSS is the indispensable source that can supply initial positioning in the commonly used high definition map-based LiDAR point cloud positioning solution for autonomous driving. However, the non-light-of-sight (NLOS) reception dominates GNSS positioning performance in super-urbanized areas. The recent proposed 3D map aided (3DMA) GNSS can mitigate the majority of the NLOS caused by buildings. However, the same phenomenon caused by moving objects in urban areas is currently not modeled in the 3D geographic information system (GIS). Therefore, we present a novel method to exclude the NLOS receptions caused by a double-decker bus, one of the symbolic tall moving objects in road transportations. To estimate the dimension and orientation of the double-decker buses relative to the GNSS receiver, LiDAR-based perception is utilized. By projecting the relative positions into GNSS Skyplot, the direct transmission path of satellite signals blocked by the moving objects can be identified and excluded from positioning. Finally, GNSS positioning is estimated by the weighted least square (WLS) method based on the remaining satellites after the NLOS exclusion. Both static and dynamic experiments are conducted in Hong Kong. The results show that the proposed NLOS exclusion using LiDAR-based perception can greatly improve the GNSS single point positioning (SPP) performance.

The recent development in vehicle-to-everything (V2X) communication opens a new opportunity to improve the positioning performance of the road users. We explore the benefit of connecting the raw data of the global navigation satellite system (GNSS) from the agents. In urban areas, GNSS positioning is highly degraded due to signal blockage and reflection. 3D building model can play a major role in mitigating the GNSS multipath and non-line-of-sight (NLOS) effects. To combine the benefits of 3D models and V2X, we propose a novel 3D mapping aided (3DMA) GNSS-based collaborative positioning method that makes use of the available surrounding GNSS receivers’ measurements. By complementarily integrating the ray-tracing based 3DMA GNSS and the double difference technique, the random errors (such as multipath and NLOS) are mitigated while eliminating the systematic errors (such as atmospheric delay and satellite clock/orbit biases) between road user. To improve the accuracy and robustness of the collaborative algorithm, factor graph optimization (FGO) is employed to optimize the positioning solutions among agents. Multiple low-cost GNSS receivers are used to collect both static and dynamic data in Hong Kong and to evaluate the proposed algorithm by post-processing. We reduce the GNSS positioning error from over 30 meters to less than 10 meters for road users in a deep urban canyon.

Mapping and localization is a critical module of autonomous driving, and significant achievements have been reached in this field. Beyond Global Navigation Satellite System (GNSS), research in point cloud registration, visual feature matching, and inertia navigation has greatly enhanced the accuracy and robustness of mapping and localization in different scenarios. However, highly urbanized scenes are still challenging: LIDAR- and camera-based methods perform poorly with numerous dynamic objects; the GNSS-based solutions experience signal loss and multi-path problems; the inertia measurement units (IMU) suffer from drifting. Unfortunately, current public datasets either do not adequately address this urban challenge or do not provide enough sensor information related to map-ping and localization. Here we present UrbanLoco: a mapping/localization dataset collected in highly-urbanized environments with a full sensor-suite. The dataset includes 13 trajectories collected in San Francisco and Hong Kong, covering a total length of over 40 kilometers. Our dataset includes a wide variety of urban terrains: urban canyons, bridges, tunnels, sharp turns, etc. More importantly, our dataset includes information from LIDAR, cameras, IMU, and GNSS receivers. Now the dataset is publicly available through the link in the footnote1.