Collision Risk and Operational Impact of Speed Change Advisories as Aircraft Collision Avoidance Maneuvers

Sydney M. Katz, Luis E. Álvarez, Michael P. Owen, Samuel S. Wu, Marc Brittain, Anshuman Das, Mykel J. KochenderferAIAA 2022 · arXiv

TL;DR

This paper investigates the effectiveness of using speed changes as a collision avoidance maneuver for future aircraft like those in Advanced Air Mobility (AAM). While speed advisories can reduce collision risk, they are less safe and efficient than traditional vertical or horizontal maneuvers, especially in challenging head-on or overtake scenarios.

What Problem Does This Paper Address?

Ensuring safe air travel relies heavily on collision avoidance systems. The FAA has developed the ACAS X family of systems, which use Markov Decision Processes (MDPs) to model and resolve potential mid-air collisions. While ACAS Xa primarily uses vertical maneuvers and ACAS Xu/sXu use horizontal maneuvers, a new variant, ACAS Xr, is being developed for rotorcraft and AAM vehicles. A key consideration for AAM is minimizing deviation from their planned flight paths. To achieve this, speed change advisories have been proposed as a less disruptive alternative to vertical or horizontal maneuvers. This paper aims to evaluate the safety and operational efficiency of these speed-based advisories, comparing them to existing ACAS X approaches.

Methods

The researchers developed an MDP-based collision avoidance logic that issues speed change advisories. This logic was then compared against existing horizontal and vertical ACAS X logics. The evaluation was conducted using Monte Carlo simulations on established airspace encounter models. They specifically analyzed the performance of speed advisories both independently and in combination with horizontal and vertical maneuvers. The safety metric used was the P(NMAC) risk ratio, which compares the probability of a Near Mid-Air Collision (NMAC) against unequipped and equipped intruders. Operational efficiency was assessed by the alert rate. They also investigated the impact of pilot response probabilities on the overall system safety.

Datasets

The paper utilizes existing encounter sets for evaluation:

  • LLCEM HMD Ext 150: Correlated flights in cruising phase, sampled from a statistical airspace model of the continental US.
  • OpSuit: Uncorrelated cruising phase encounters, paired with intruder tracks from other studies.
  • Hovering: Uncorrelated hovering vs. cruising encounters, designed to test scenarios with hovering ownships.

These encounter sets are standard benchmarks used in the evaluation of ACAS X variants.

Key Insights and Ideas

  • Speed is a compromise: While speed changes can help avoid collisions, they come with significant trade-offs in safety and efficiency compared to traditional maneuvers.
  • Trajectory preservation has limits: The desire to stick to a prescribed flight path for AAM vehicles is a strong motivator for speed advisories, but this benefit is outweighed by safety concerns in certain critical scenarios.
  • Head-on and overtake are speed's Achilles' heel: Speed advisories are particularly ineffective in resolving head-on or co-altitude overtake situations, where they can only delay, not prevent, a collision.
  • Blended logics increase alerts: Combining speed advisories with horizontal or vertical maneuvers improves safety slightly but leads to a substantial increase in alerts.
  • Pilot response is crucial: For human pilots, the added complexity of a third advisory dimension (speed) can actually decrease overall system safety if pilots struggle to respond to all advisories simultaneously.

Key Results

  • Speed-only logic is significantly less safe: The speed-only logic exhibited P(NMAC) risk ratios 6 to 69 times higher in the LLCEM encounter set and alerted 3 to 4 times more frequently in the OpSuit set compared to horizontal and vertical logics.
  • Speed logic struggles with ICAO targets: The speed logic alone did not meet the International Civil Aviation Organization (ICAO) target risk ratios for collision avoidance systems.
  • Hovering encounters highlight weaknesses: In hovering encounters, the speed logic had a significantly higher unequipped risk ratio (0.209) compared to vertical (0.028) and horizontal (0.172) logics.
  • Blended logics are safer but noisier: When speed advisories were combined with horizontal or vertical logics, the risk ratio decreased, but the alert rate increased significantly.
  • Pilot response can degrade safety: Introducing speed advisories, when considering the probability of a pilot not responding to an advisory dimension, can increase the overall system P(NMAC) by 44% (from 3.31 x 10⁻⁴ to 4.77 x 10⁻⁴).

Limitations and Future Work

  • Pilot response modeling: The paper acknowledges that the pilot response model is a simplification. Future work could explore more sophisticated models of pilot behavior and decision-making.
  • Automated systems: The authors suggest that for systems with automated responses, where pilot response is not a factor, speed advisories might offer a safety benefit. This indicates a direction for future research in fully autonomous collision avoidance.
  • Operational cost: The increased alert rate associated with speed advisories is a significant operational cost that needs further consideration.
  • Specific AAM capabilities: While the speed advisories were designed with AAM in mind, further tuning and validation with specific AAM vehicle performance characteristics could be beneficial.

Summary updated

Abstract

View Video Presentation: https://doi.org/10.2514/6.2022-3824.vid Aircraft collision avoidance systems have long been a key factor in keeping our airspace safe. Over the past decade, the FAA has supported the development of a new family of collision avoidance systems called the Airborne Collision Avoidance System X (ACAS X), which model the collision avoidance problem as a Markov decision process (MDP). Variants of ACAS X have been created for both manned (ACAS Xa) and unmanned aircraft (ACAS Xu and ACAS sXu). The variants primarily differ in the types of collision avoidance maneuvers they issue. For example, ACAS Xa issues vertical collision avoidance advisories, while ACAS Xu and ACAS sXu allow for horizontal advisories due to reduced aircraft performance capabilities. Currently, a new variant of ACAS X, called ACAS Xr, is being developed to provide collision avoidance capability to rotorcraft and Advanced Air Mobility (AAM) vehicles. Due to the desire to minimize deviation from the prescribed flight path of these aircraft, speed adjustments have been proposed as a potential collision avoidance maneuver for aircraft using ACAS Xr. In this work, we investigate the effect of speed change advisories on the safety and operational efficiency of collision avoidance systems. We develop an MDP-based collision avoidance logic that issues speed advisories and compare its performance to that of horizontal and vertical logics through Monte Carlo simulation on existing airspace encounter models. Our results show that while speed advisories are able to reduce collision risk, they are neither as safe nor as efficient as their horizontal and vertical counterparts.
BibTeX
@misc{katz2022collision,
  author = {Sydney M. Katz and Luis E. Álvarez and Michael P. Owen and Samuel S. Wu and Marc Brittain and Anshuman Das and Mykel J. Kochenderfer},
  title = {{Collision Risk and Operational Impact of Speed Change Advisories as Aircraft Collision Avoidance Maneuvers}},
  year = {2022},
  eprint = {2204.14250},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2204.14250},
}
Figure