Scalable Multi-Agent Computational Guidance with Separation Assurance for Autonomous Urban Air Mobility

Yang, Xuxi; Wei, Peng

doi:10.2514/1.g005000

Cited by 68 publications

(33 citation statements)

References 56 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Besides public transportation discussed above, there are also some other application scenarios for flying cars in which transportation requirements are dynamic, such as shortdistance travel and sightseeing, and customized personalized services. Operators of flying cars should be able to quickly offer the customers convenient and efficient path planning services according to their on-demand requirements [52]- [54].…”

Section: ) Adaptive Path Planningmentioning

confidence: 99%

Flying Car Transportation System: Advances, Techniques, and Challenges

Pan

Alouini

2021

IEEE Access

View full text Add to dashboard Cite

Since the development of transport systems, humans have exploited ground-level, belowground, and high-altitude spaces for transportation purposes. However, with the increasing burden of expanding populations and rapid urbanization in recent decades, public transportation systems and freight traffic are suffering huge pressure, plaguing local governments and straining economies. Engineers and researchers have started to re-examine, propose, and develop the underused near-ground spaces (NGS) for transportation purposes. For instance, flying cars, which are not a totally novel idea, aim at solving the traffic congestion problem and releasing the strains on existing city transport networks by utilizing unoccupied NGS. Flying cars differ from traditional grounded transportation systems that are entirely limited by their physical space, such as trains on tracks or automobiles on roads. Flying cars do not occupy or compete for high-altitude spaces used by air traffic for long-distance transfer. As there is limited specific literature on flying cars and flying car transportation systems (FCTS), this paper aims to describe the modern advances, techniques, and challenges of FCTS. We explore the inherent nature of NGS transportation and devise useful proposals to facilitate the construction and commercialization of FCTS. We begin with an introduction to the increasing need for NGS transportation and we address the advantages of using flying cars. Next, we present a brief overview of the history of the development of flying cars in terms of the historic timeline and technique development. Then, we discuss and compare the state of the art in the design of flying cars, including the take-off & landing (TOL) modes, pilot modes, operation modes, and power types, which are related to the adaptability, flexibility & comfort, stability & complexity, and environmental friendliness of flying cars, respectively. Additionally, since large-scale operations of flying cars can improve current transportation problems, we also introduce different facets of the various designs of FCTS, including path and trajectory planning, supporting facilities, and commercial designs. Finally, we discuss the challenges that might arise while developing and commercializing FCTS in terms of safety issues, commercial issues, and ethical issues.

show abstract

Section: ) Adaptive Path Planningmentioning

confidence: 99%

Flying Car Transportation System: Advances, Techniques, and Challenges

Pan

Alouini

2021

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Although the above algorithms perform well in specific scenarios, they cannot adapt to stochastic dynamic models. Yang and Wei presented a message-based decentralized computational guidance algorithm which is implemented by the Monte Carlo tree search technique [13].…”

Section: Introductionmentioning

confidence: 99%

Reinforcement Learning‐Based Collision Avoidance Guidance Algorithm for Fixed‐Wing UAVs

et al. 2021

View full text Add to dashboard Cite

A deep reinforcement learning-based computational guidance method is presented, which is used to identify and resolve the problem of collision avoidance for a variable number of fixed-wing UAVs in limited airspace. The cooperative guidance process is first analyzed for multiple aircraft by formulating flight scenarios using multiagent Markov game theory and solving it by machine learning algorithm. Furthermore, a self-learning framework is established by using the actor-critic model, which is proposed to train collision avoidance decision-making neural networks. To achieve higher scalability, the neural network is customized to incorporate long short-term memory networks, and a coordination strategy is given. Additionally, a simulator suitable for multiagent high-density route scene is designed for validation, in which all UAVs run the proposed algorithm onboard. Simulated experiment results from several case studies show that the real-time guidance algorithm can reduce the collision probability of multiple UAVs in flight effectively even with a large number of aircraft.

show abstract

“…While MDP methods for collision avoidance can be solved offline in the pre-depature phase Chryssanthacopoulos and Kochenderfer, 2012a,b;, recent work has shown that MDPs can be efficiently solved online in-flight (Yang and Wei, 2020;. Online methods are more adaptable to changes in environment given the ability to sense new states and react accordingly, whereas the policy for offline methods is computed prior to take-off.…”

Section: Chapter 1 Introductionmentioning

confidence: 99%

“…during flight . MDP based methods can be solved offline during the pre-departure phase Chryssanthacopoulos and Kochenderfer, 2012a,b; or online during the en route phase (Yang and Wei, 2020;. Offline methods are typically not adaptive to changes in the environment because the policy is determined ahead of time.…”

Section: Introductionmentioning

confidence: 99%

“…Through offline training, deep reinforcement learning algorithms can use an approximation function to represent the policy over a continuous state-action space without space discretization. In addition, the execution of deep reinforcement learning algorithms is very efficient compared to approaches that solve the MDP online (Yang and Wei, 2020;. The first work in using reinforcement learning in a tactical air traffic control and conflict resolution application is described in .…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Learning-based decision making for safe and scalable autonomous separation assurance

Brittain¹

View full text Add to dashboard Cite

A novel deep multi-agent reinforcement learning framework is proposed to identify and resolve conflicts among a variable number of aircraft in a high-density, stochastic, and dynamic en route sector.The concept of using distributed vehicle autonomy to ensure separation is proposed, instead of a centralized sector air traffic controller. Our proposed framework utilizes Proximal Policy Optimization (PPO) that is customized to incorporate an attention network. This allows the agents to have access to variable aircraft information in the sector in a scalable, efficient approach to achieve high traffic throughput under uncertainty. Agents are trained using a centralized learning, decentralized execution scheme where one neural network is learned and shared by all agents. The proposed framework is validated on three case studies in the BlueSky air traffic simulator. Several baselines are introduced and the numerical results show that the proposed framework significantly reduces offline training time, increases safe separation performance, and results in a more efficient policy.

show abstract

Scalable Multi-Agent Computational Guidance with Separation Assurance for Autonomous Urban Air Mobility

Cited by 68 publications

References 56 publications

Flying Car Transportation System: Advances, Techniques, and Challenges

Flying Car Transportation System: Advances, Techniques, and Challenges

Reinforcement Learning‐Based Collision Avoidance Guidance Algorithm for Fixed‐Wing UAVs

Learning-based decision making for safe and scalable autonomous separation assurance

Contact Info

Product

Resources

About