scispace - formally typeset
Search or ask a question

Showing papers on "Collision avoidance system published in 2019"


Journal ArticleDOI
TL;DR: This study proposes an efficient method to overcome multiship collision avoidance problems based on the Deep Reinforcement Learning (DRL) algorithm and demonstrates its excellent adaptability to unknown complex environments with various encountered ships.

127 citations


Journal ArticleDOI
TL;DR: The proposed GVO-CAS can offer rule-compliant evasive actions with a minimum number of required actions for ships and shows the great potential to use the GVO algorithm in both manned and unmanned ships at sea.

124 citations


Journal ArticleDOI
TL;DR: A novel approach based on deep reinforcement learning (DRL) is proposed for automatic collision avoidance of multiple ships particularly in restricted waters, incorporating ship manoeuvrability, human experience and navigation rules.

121 citations


Proceedings ArticleDOI
02 May 2019
TL;DR: An assistive suitcase system for supporting blind people when walking through crowded environments using pre-emptive sound notifications, BBeep, and it is observed that the proposed system significantly reduces the number of imminent collisions.
Abstract: We present an assistive suitcase system, BBeep, for supporting blind people when walking through crowded environments. BBeep uses pre-emptive sound notifications to help clear a path by alerting both the user and nearby pedestrians about the potential risk of collision. BBeep triggers notifications by tracking pedestrians, predicting their future position in real-time, and provides sound notifications only when it anticipates a future collision. We investigate how different types and timings of sound affect nearby pedestrian behavior. In our experiments, we found that sound emission timing has a significant impact on nearby pedestrian trajectories when compared to different sound types. Based on these findings, we performed a real-world user study at an international airport, where blind participants navigated with the suitcase in crowded areas. We observed that the proposed system significantly reduces the number of imminent collisions.

100 citations


Proceedings ArticleDOI
15 Dec 2019
TL;DR: A method to provide safety guarantees when using a neural network collision avoidance system is proposed and experiments with systems inspired by ACAS X show that neural networks giving either horizontal or vertical maneuvers can be proven safe.
Abstract: The decision logic for the ACAS X family of aircraft collision avoidance systems is represented as a large numeric table. Due to storage constraints of certified avionics hardware, neural networks have been suggested as a way to significantly compress the data while still preserving performance in terms of safety. However, neural networks are complex continuous functions with outputs that are difficult to predict. Because simulations evaluate only a finite number of encounters, simulations are not sufficient to guarantee that the neural network will perform correctly in all possible situations. We propose a method to provide safety guarantees when using a neural network collision avoidance system. The neural network outputs are bounded using neural network verification tools like Reluplex and Reluval, and a reachability method determines all possible ways aircraft encounters will resolve using neural network advisories and assuming bounded aircraft dynamics. Experiments with systems inspired by ACAS X show that neural networks giving either horizontal or vertical maneuvers can be proven safe. We explore how relaxing the bounds on aircraft dynamics can lead to potentially unsafe encounters and demonstrate how neural network controllers can be modified to guarantee safety through online costs or lowering alerting cost. The reachability method is flexible and can incorporate uncertainties such as pilot delay and sensor error. These results suggest a method for certifying neural network collision avoidance systems for use in real aircraft.

44 citations


Journal ArticleDOI
TL;DR: This paper proposed a secure and privacy-preserving collision avoidance system in 5G fog based IoV that makes use of certificateless aggregate signcryption coupled with pseudonymous technique as the building blocks to ensure authentication, integrity, confidentiality and privacy preservation.

39 citations


Journal ArticleDOI
TL;DR: A novel collaborative driving scheme is designed by fusing the control inputs from the human driver and the co-pilot to obtain the final control input for the ego vehicle under different circumstances.
Abstract: With a goal to improve transportation safety, this paper proposes a collaborative driving framework based on assessments of both internal and external risks involved in vehicle driving. The internal risk analysis includes driver drowsiness detection and driver intention recognition that helps to understand the human driver’s behavior. Steering wheel data and facial expression are used to detect the driver’s drowsiness. Hidden Markov models are adapted to recognize the driver’s intention using the vehicle’s lane position, control, and state data. For the external risk analysis, a co-pilot utilizes a collision avoidance system to estimate the collision probability between the ego vehicle and other nearby vehicles. Based on the risk analyses, we design a novel collaborative driving scheme by fusing the control inputs from the human driver and the co-pilot to obtain the final control input for the ego vehicle under different circumstances. The proposed collaborative driving framework is validated in an assisted-driving testbed, which enables both autonomous and manual driving capabilities.

39 citations


Journal ArticleDOI
TL;DR: The simulation results conclusively demonstrate that this method can achieve collision avoidance in the presence of an unknown acceleration of obstacle, and the performance of collision avoidance using this method is greater than the collision avoidance method based on normal finite time convergent guidance.

38 citations


Proceedings ArticleDOI
01 Sep 2019
TL;DR: ACAS sXu provides collision avoidance systems development over the past decade, enabling autonomous and decentralized sUAS collision avoidance capability against manned aircraft, UAS, and other s UAS.
Abstract: Demand for small unmanned aircraft systems (sUAS) continues to increase in diversity and volume, however applications are currently limited by regulatory requirements for visual observation or special use waivers. For full integration of autonomous sUAS in the national airspace system, a collision avoidance system must be implemented to enable detection and avoidance of air traffic. Building upon collision avoidance systems development over the past decade for large UAS and manned aircraft, ACAS sXu provides such a capability, enabling autonomous and decentralized sUAS collision avoidance capability against manned aircraft, UAS, and other sUAS.

28 citations


Journal ArticleDOI
TL;DR: A method based on finite control set model predictive control can control formation quickly to avoid obstacles and reach the destination in accordance with the dynamics of each vessel in the formation, without the prior knowledge of the environment and reference trajectory.
Abstract: This paper deals with the problem of formation collision avoidance for unmanned surface vehicles (USVs). Compared with the generalship formation, the formation collision avoidance system (FCAS) needs better responsiveness and stability because of faster speed and smaller volume for USVs. A method based on finite control set model predictive control is proposed to solve this problem. The novelty of the method is that it can control formation quickly to avoid obstacles and reach the destination in accordance with the dynamics of each vessel in the formation, without the prior knowledge of the environment and reference trajectory. The thruster speed and propulsion angle of the USV form a finite control set, which is more practical. The FCAS adopts the leader–follower structure and distributed control strategy to ensure that the followers have a certain autonomy. The first two simulation tests verify that the system has the formation stability, formation forming ability, and the applicability in restricted water. The last simulation test shows that the system can control the USV formation to sail quickly and safely in complex sea scenarios with formation transformation task and multiple dynamic obstacles.

27 citations


Posted Content
TL;DR: The results show that a corrected collision avoidance system can operate more efficiently than traditional methods in dense airspace while maintaining high levels of safety.
Abstract: New methodologies will be needed to ensure the airspace remains safe and efficient as traffic densities rise to accommodate new unmanned operations. This paper explores how unmanned free-flight traffic may operate in dense airspace. We develop and analyze autonomous collision avoidance systems for aircraft operating in dense airspace where traditional collision avoidance systems fail. We propose a metric for quantifying the decision burden on a collision avoidance system as well as a metric for measuring the impact of the collision avoidance system on airspace. We use deep reinforcement learning to compute corrections for an existing collision avoidance approach to account for dense airspace. The results show that a corrected collision avoidance system can operate more efficiently than traditional methods in dense airspace while maintaining high levels of safety.

Proceedings ArticleDOI
09 Jun 2019
TL;DR: In this paper, a partially observable Markov decision process (POMDP) is used to derive a policy robust to uncertainty in the pedestrian location, which is then integrated with an AEB system that operates only when a collision is unavoidable.
Abstract: Safe autonomous driving in urban areas requires robust algorithms to avoid collisions with other traffic participants with limited perception ability. Current deployed approaches relying on Autonomous Emergency Braking (AEB) systems are often overly conservative. In this work, we formulate the problem as a partially observable Markov decision process (POMDP), to derive a policy robust to uncertainty in the pedestrian location. We investigate how to integrate such a policy with an AEB system that operates only when a collision is unavoidable. In addition, we propose a rigorous evaluation methodology on a set of well-defined scenarios. We show that combining the two approaches provides a robust autonomous braking system that reduces unnecessary braking caused by using the AEB system on its own.

Proceedings ArticleDOI
01 Oct 2019
TL;DR: A collision avoidance algorithm designed for a robotic redundant task is presented that permits the robot to avoid collisions with obstacles and in the same time to achieve the planned task when possible.
Abstract: Collaborative robotics implies the possibility for robots and humans to cooperate to perform a common task. A fundamental element to make collaboration possible is to ensure the safety of the human operators. In this paper a collision avoidance algorithm designed for a robotic redundant task is presented. This algorithm permits the robot to avoid collisions with obstacles (e.g. human operators) and in the same time to achieve the planned task when possible. The algorithm was tested with a UR3 robot of Universal Robots company, considering only tool centre point (TCP) position in order to obtain kinematic redundancy. The robot is controlled by an external PC that communicates with the controller of the robot via Ethernet network. The communication and control algorithms are written in the Matlab environment. The PC-UR communication functions permit to send joint velocities commands from the PC to the controller of the robot and feedback data from the UR to the PC. Control functions estimate the distances between an obstacle (whose position was acquired by a system of motion capture) and the robot and calculate the needed joints velocities to avoid any collision. The functions developed in Matlab will be explained in detail. Results of the experimental tests conducted to verify the effectiveness of the collision avoidance system are reported.

Proceedings ArticleDOI
01 Oct 2019
TL;DR: A collision avoidance system based on vision sensors that is suitable for collaborative robotics scenarios and the information of the position of a human operator is obtained by Microsoft Kinect sensor in the form of a point cloud.
Abstract: This paper presents a collision avoidance system based on vision sensors that is suitable for collaborative robotics scenarios. In fact, collaborative robotics foresees the possibility that humans and robots share the same workspace, so the safety of the human operators must be ensured. The collision avoidance algorithm here presented can modify the trajectory of the robot in order to avoid any collisions with a human operator. A fundamental element for the algorithm is the relative position between robot and human. In this work, the information of the position of a human operator is obtained by Microsoft Kinect sensor in the form of a point cloud. Two Microsoft Kinect are used and their point cloud data is merged to overcome the problems related to the possible occlusions of the sensors, obtaining a more reliable point cloud. Each Kinect works with a dedicated PC and the two PCs communicate via ethernet network in a master-slave mode. The layout of the acquiring system is described and the functions used for the communication between the PCs and the manipulation of the point clouds are presented. Results of simulation tests made to verify the performances of the system and collision avoidance based on point cloud compared with convex mesh are reported and discussed.

Journal ArticleDOI
TL;DR: The extended Kalman filter (EKF) scheme that combines UWB with INS data is used to improve the localization accuracy of AGV and experiments are given to show that the EKF scheme can get accurate position estimation and the collisions among AGVs can be detected and avoided in time.
Abstract: As a highly automated carrying vehicle, an automated guided vehicle (AGV) has been widely applied in various industrial areas. The collision avoidance of AGV is always a problem in factories. Current solutions such as inertial and laser guiding have low flexibility and high environmental requirements. An INS (inertial navigation system)-UWB (ultra-wide band) based AGV collision avoidance system is introduced to improve the safety and flexibility of AGV in factories. An electronic map of the factory is established and the UWB anchor nodes are deployed in order to realize an accurate positioning. The extended Kalman filter (EKF) scheme that combines UWB with INS data is used to improve the localization accuracy. The current location of AGV and its motion state data are used to predict its next position, decrease the effect of control delay of AGV and avoid collisions among AGVs. Finally, experiments are given to show that the EKF scheme can get accurate position estimation and the collisions among AGVs can be detected and avoided in time.

Patent
04 Jul 2019
TL;DR: In this paper, the authors present methods, devices, and robotic vehicles that adjust a proximity threshold implemented in a collision avoidance system based on whether a payload is being carried by the robotic vehicle.
Abstract: Various embodiments include methods, devices, and robotic vehicles that adjust a proximity threshold implemented in a collision avoidance system based on whether a payload is being carried. Methods may include determining whether a payload is carried by the robotic vehicle, setting a proximity threshold for collision avoidance in response to determining that a payload is carried by the robotic vehicle, and controlling one or more motors of the robotic vehicle using the proximity threshold for collision avoidance. Some embodiments may include raising the proximity threshold when a payload is not being carried or decreasing proximity threshold when a payload is being carried. Some embodiments may include determining a classification of a payload and setting the proximity threshold based at least in part on the classification.

30 Oct 2019
TL;DR: Zinchenko, S., Nosov, P., Mateichuk, V., Mamenko, P, Popovych, I. & Grosheva, O. as mentioned in this paper proposed an automatic collision avoidance system with many targets, including maneuvering ones.
Abstract: Zinchenko, S., Nosov, P., Mateichuk, V., Mamenko, P., Popovych, I. & Grosheva, O. (2019). Automatic collision avoidance system with many targets, including maneuvering ones. Bulletin of university of Karaganda, 96 (4), 69-79. DOI: 0.31489/2019Ph4/69-79

Proceedings ArticleDOI
29 Apr 2019
TL;DR: This paper proposes a MEC-based cooperative Collision Avoidance (MECAV) system designed to anticipate the detection and localization of road hazards, and has implemented a proof-of-concept of the system and validated its architecture and functionalities.
Abstract: Mobile Edge Computing (MEC) is a key enabler for the deployment of vehicular use cases, as it guarantees low latency and high bandwidth requirements. In this paper, we propose a MEC-based cooperative Collision Avoidance (MECAV) system designed to anticipate the detection and localization of road hazards. It includes a Collision Avoidance service, allocated in the MEC infrastructure, which receives information of status and detected hazards from vehicles, processes this information, and selectively informs to each vehicle that either approaches to road hazards or to other vehicles. We have implemented a proof-of-concept of the MECAV system and we have validated its architecture and functionalities.

Journal ArticleDOI
TL;DR: A parallel trajectory planning architecture is proposed in this paper for SACAS system to ensure the collision-free optimal navigation in compliance with COLREGs rules.

Journal ArticleDOI
TL;DR: The results showed that female drivers were more likely to be involved with RLR collisions, and male drivers could detect the conflicting RLR vehicle more quickly than female drivers, which could direct warning condition design to improve the effectiveness of collision avoidance systems.
Abstract: Intersections have been recognized as hazard locations with lots of visual information that drivers need to process. Although the collision avoidance systems (CASs) have been proved to effectively reduce the crash rate and much research on the effectiveness of CASs has been conducted with regard to the driving behaviors, drivers’ visual performances under the effects of different collision avoidance warning conditions that were closely related to the effectiveness of CAS have been neglected. In this study, a driving simulator experiment was conducted to evaluate the relationships among drivers’ visual performances, drivers’ different warning conditions (warning timings warning content) and driver’s gender when they crossed the intersections involved with red-light running (RLR) vehicles. The experimental results showed that warning timings had significant effects on the detection stage and reaction stage. Specifically, drivers could detect the conflicting RLR vehicle most quickly in the warning timings of 4.5 s ahead of a collision. When the warning was released earlier than 5.0 s ahead of a collision, driver tended to take brake action earlier than paying a fixation on a conflicting RLR vehicle. Warning content only had significant effects on drivers’ detection stage. Compared to the non-directional warning, the specific directional information could shorten the time spent in detecting the conflicting RLR vehicle. Besides, directional information could increase drivers’ average blink duration during the process of collision avoidance. Additionally, the results showed that female drivers were more likely to be involved with RLR collisions, and male drivers could detect the conflicting RLR vehicle more quickly than female drivers. Also, it had been found that later warning timings tended to increase female drivers’ blink rate, and non-directional warning tended to increase female drivers’ blink rate. These findings could direct warning condition design to improve the effectiveness of collision avoidance systems.

Posted Content
TL;DR: It is found that, thanks to MEC, the protection of collision avoidance, traditionally thought for vehicles, is able to be extended to vulnerable users without impacting its effectiveness or latency.
Abstract: Collision avoidance is one of the most promising applications for vehicular networks, dramatically improving the safety of the vehicles that support it. In this paper, we investigate how it can be extended to benefit vulnerable users, e.g., pedestrians and bicycles, equipped with a smartphone. We argue that, owing to the reduced capabilities of smartphones compared to vehicular on-board units, traditional distributed approaches are not viable, and that multi-access edge computing (MEC) support is needed. Thus, we propose a MEC-based collision avoidance system, discussing its architecture and evaluating its performance. We find that, thanks to MEC, we are able to extend the protection of collision avoidance, traditionally thought for vehicles, to vulnerable users without impacting its effectiveness or latency.

Journal ArticleDOI
Yu Kai, Peng Liqun, Ding Xue, Fan Zhang, Minrui Chen 
10 Jun 2019
TL;DR: This study proposes a data fusion method to capture the vehicle crash risk by extracting critical information from raw BSMs data, such as driver volition, vehicle speed, hard accelerations and braking, and a classification model based on information-entropy and variable precision rough set is used.
Abstract: Basic safety message (BSM) is a core subset of standard protocols for connected vehicle system to transmit related safety information via vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I). Although some safety prototypes of connected vehicle have been proposed with effective strategies, few of them are fully evaluated in terms of the significance of BSM messages on performance of safety applications when in emergency.,To address this problem, a data fusion method is proposed to capture the vehicle crash risk by extracting critical information from raw BSMs data, such as driver volition, vehicle speed, hard accelerations and braking. Thereafter, a classification model based on information-entropy and variable precision rough set (VPRS) is used for assessing the instantaneous driving safety by fusing the BSMs data from field test, and predicting the vehicle crash risk level with the driver emergency maneuvers in the next short term.,The findings and implications are discussed for developing an improved warning and driving assistant system by using BSMs messages.,The findings of this study are relevant to incorporation of alerts, warnings and control assists in V2V applications of connected vehicles. Such applications can help drivers identify situations where surrounding drivers are volatile, and they may avoid dangers by taking defensive actions.

01 Jan 2019
TL;DR: In this paper, the authors formulate the problem as a partially observable Markov decision process (POMDP) to derive a policy robust to uncertainty in the pedestrian location and investigate how to integrate such a policy with an AEB system that operates only when a collision is unavoidable.
Abstract: Safe autonomous driving in urban areas requires robust algorithms to avoid collisions with other traffic participants with limited perception ability. Current deployed approaches relying on Autonomous Emergency Braking (AEB) systems are often overly conservative. In this work, we formulate the problem as a partially observable Markov decision process (POMDP), to derive a policy robust to uncertainty in the pedestrian location. We investigate how to integrate such a policy with an AEB system that operates only when a collision is unavoidable. In addition, we propose a rigorous evaluation methodology on a set of well defined scenarios. We show that combining the two approaches provides a robust autonomous braking system that reduces unnecessary braking caused by using the AEB system on its own.

21 Mar 2019
TL;DR: A multi-path automatic ground collision avoidance system (Auto-GCAS) for performance limited aircraft was further developed and improved to prevent controlled flight into terrain.
Abstract: A multi-path automatic ground collision avoidance system (Auto-GCAS) for performance limited aircraft was further developed and improved to prevent controlled flight into terrain. This research includes flight test results from the United States Test Pilot School’s Test Management Project (TMP) titled Have Multi-Path Escape Decisions Using Sophisticated Algorithms (MEDUSA). Currently, the bomber and mobility aircraft communities lack an Auto-GCAS. The F-16 Auto-GCAS was proven successful for fighter-type aircraft with seven aircraft and eight lives saved from 2014 to 2018. The newly developed and tested Rapidly Selectable Escape Trajectory (RSET) system included a 5-path implementation which continuously updated at a rate of up to 12.5 Hz. The research employed Level 1 Digital Terrain Elevation Data (DTED) to identify the offending terrain and an augmented 6 Degree-of-Freedom (DoF) Stitched aerodynamic model to create terrain avoidance paths based on the aircraft’s current state and location. The system then triggered when all paths predicted collision with the DTED and automatically activated the path which had the longest time until impact. A terrain safety buffer (TSB) of 200 ft added to the DTED to allowed for the time needed to process and execute the maneuver. The RSET system was flight tested against DTED using the Calspan Learjet 25D Variable Stability System (VSS). Path prediction error (PPE) did not meet the specified criteria and was larger than expected for the 30-second path predictions; however, at the maximum refresh rate of 12.5 Hz, the RSET system ensured terrain clearance in all cases tested. The RSET system was able to achieve and maintain target load factor and flight path angle with momentary overshoots. The system showed no tendency for nuisance. The RSET hand-back was favorable and can be used as a baseline for future Auto-GCASs.

Journal ArticleDOI
01 Nov 2019
TL;DR: A new solution based on smartphones carried by drivers and pedestrians is proposed so that it is the device inside the vehicle violating a traffic light that self-reports the offence in order to generate alerts and warn nearby vehicles and pedestrians to prevent accidents.
Abstract: In this paper, a collision avoidance system is presented to detect red light running and warn nearby vehicles and pedestrians in real time in order to prevent possible accidents. No complex infrastructure-based solution such as those based on radars or cameras is here required. Instead, a new solution based on smartphones carried by drivers and pedestrians is proposed so that it is the device inside the vehicle violating a traffic light, the one that self-reports the offence in order to generate alerts and warn nearby vehicles and pedestrians to prevent accidents. The proposal could also be used by road authorities to collect data on traffic lights that are most frequently violated in order to define an action plan to investigate causes and look for solutions. It includes a classifier for learning and estimating driver behaviour based on collected data, which is used to predict whether he/she is about to run a red light or detect whether that has already happened. In the first case, the system broadcasts warnings directly to close vehicles and pedestrians through Wi-Fi, while in the second case, the proposal warns vehicles and pedestrians in the neighbourhood through a server. The solution also includes a prioritization system based on changing traffic lights at intersections according to the needs and characteristics of the traffic at all times, giving the top priority to emergency vehicles. Furthermore, the proposal involves the use of cryptographic schemes to protect authenticity and integrity of messages sent from traffic lights, smartphones and servers, and privacy and anonymity to promote the use of the system. A beta version with some parts of the proposal has been implemented and the obtained results are promising.


Patent
25 Jan 2019
TL;DR: In this article, a ship intelligent collision avoidance system based on maneuverrability modeling, comprising a state sensing subsystem, where the state parameters of the ship and the position information of an obstacle are obtained; the maneuverability modeling module processes the ship's own state parameters, constructs the sample pairs, carries on the ship maneuverability on-line modeling, and predicts the ship possible arrival position at the next time under all feasible maneuverability.
Abstract: The invention provides a ship intelligent collision avoidance system based on manoeuvrability modeling, comprising a state sensing subsystem, wherein the state parameters of the ship and the positioninformation of an obstacle are obtained; the maneuverability modeling module processes the ship's own state parameters, constructs the sample pairs, carries on the ship maneuverability on-line modeling, and predicts the ship's possible arrival position at the next time under all feasible maneuverability. The intelligent collision avoidance module combines the position information of the obstacle,Binary navigable area information and collision avoidance rules are used to carry out dynamic path planning. In path planning, the maneuverability modeling module predicts the possible arrival position of the ship at the next time as a constraint, outputs a reasonable planning path point sequence, and decouples it into a heading tracking sequence and a speed tracking sequence. Track the planned real-time heading and speed respectively. The invention realizes the intelligent collision avoidance decision of the ship on the basis of the on-line prediction of the ship maneuverability, and realizesthe safe and autonomous navigation of the ship.

Patent
04 Apr 2019
TL;DR: In this article, a collision warning system determines probabilities of potential collisions between a vehicle and other objects such as other vehicles by tracking motion of the object, and provides a notification of the potential collision to a driver of the vehicle.
Abstract: A collision warning system determines probabilities of potential collisions between a vehicle and other objects such as other vehicles. In an embodiment, sensors of a client device capture sensor data including motion data and image frames from a forward-facing view of the vehicle. An orientation of the client device relative to the vehicle may be determined using the motion data. The collision warning system determines cropped portions of the image frames and detects an object captured the image frames by processing the cropped portions. The collision warning system determines a probability of a potential collision between the vehicle and the object by tracking motion of the object. Responsive to determining that the probability is greater than a threshold value, the collision warning system may provide a notification of the potential collision to a driver of the vehicle.

Book ChapterDOI
24 May 2019
TL;DR: In this paper, an LGMD based collision avoidance method for UAV indoor navigation is presented, where four individual competitive LGMDs compete for guiding the directional collision avoidance of UAV.
Abstract: Building a reliable and efficient collision avoidance system for unmanned aerial vehicles (UAVs) is still a challenging problem. This research takes inspiration from locusts, which can fly in dense swarms for hundreds of miles without collision. In the locust’s brain, a visual pathway of LGMD-DCMD (lobula giant movement detector and descending contra-lateral motion detector) has been identified as collision perception system guiding fast collision avoidance for locusts, which is ideal for designing artificial vision systems. However, there is very few works investigating its potential in real-world UAV applications. In this paper, we present an LGMD based competitive collision avoidance method for UAV indoor navigation. Compared to previous works, we divided the UAV’s field of view into four subfields each handled by an LGMD neuron. Therefore, four individual competitive LGMDs (C-LGMD) compete for guiding the directional collision avoidance of UAV. With more degrees of freedom compared to ground robots and vehicles, the UAV can escape from collision along four cardinal directions (e.g. the object approaching from the left-side triggers a rightward shifting of the UAV). Our proposed method has been validated by both simulations and real-time quadcopter arena experiments.

Proceedings ArticleDOI
14 Jul 2019
TL;DR: An approach of real-time collision avoidance has been presented with the compliance with the COLREGS rules been successfully integrated for USV, designed in a way that through the judgment of the collision situation, the velocity and heading angle of the USV are changed to complete the collision avoidance of the obstacle.
Abstract: Unmanned surface vehicles (USVs) are becoming increasingly vital in a variety of maritime applications. The development of a real-time autonomous collision avoidance system is the pivotal issue in the study on USVs, in which the reliable collision risk detection and the adoption of a plausible collision avoidance maneuver play a key role. Existing studies on this subject seldom integrate the International Regulations for Preventing Collisions at Sea 1972 (COLREGS) guidelines. However, in order to ensure maritime safety, it is of fundamental importance that such a regulation should be obeyed at all times. In this paper, an approach of real-time collision avoidance has been presented with the compliance with the COLREGS rules been successfully integrated for USV. The approach has been designed in a way that through the judgment of the collision situation, the velocity and heading angle of the USV are changed to complete the collision avoidance of the obstacle. A strategy with reference obstacle is proposed to deal with the multiple moving obstacles situation. A number of simulations have been conducted in order to confirm the validity of the theoretic results obtained. The results show that the algorithms can sufficiently deal with complex traffic environments and that the generated practical path is suitable for USVs.