top of page


  • Writer's pictureRefMap

RefMap project participated in the Aerospace Europe Conference 2023 in Lausanne

The Aerospace Europe Conference is an event that brings cutting-edge ideas about the future of aviation. This year it took place on 9th and 10th July in the Swiss city of Lausanne. 2023’s conference was particularly significant as it united two prestigious events - the 10th European Conference for Aerospace Sciences (EUCASS) and the 9th conference of the Council of European Aerospace Societies (CEAS). Bringing together experts, researchers, and industry leaders, the conference provided a platform to exchange ideas and innovations that aim to shape the future of aviation.

The RefMap team from UC3M

Among the participants were members of the RefMap team from Carlos III University of Madrid (UC3M), whose research work earned them the opportunity to present two conference papers. These papers explored innovative advancements in the trajectory optimization for minimum environmental impact, aiming to foster a more sustainable and climate-resilient aviation industry.

1. Robust Climate Optimal Flight Planning: Identifying "Win-Win" Solutions

The first conference paper, presented by Abolfazl Simorgh, focused on pointing out climate-optimal flight planning solutions that are also beneficial from an economic point of view: despite being a promising measure to mitigate non-CO2 climate effects, climate-optimal flight planning is generally not cost-effective. However, aviation stakeholders can address this issue by implementing market-based instruments, which are currently lacking for non-CO2 emissions. The research study proposes a novel flight planning framework, which accounts for the cost of non-CO2 emissions alongside CO2 emissions in order to plan trajectories that mitigate both operational costs and climate effects. The costs of non-CO2 species are quantified within the concept of equivalent CO2 emissions. The prototype algorithmic climate change function is used to convert non-CO2 emissions as equivalent CO2 emissions by using average temperature response as a climate metric. This selection allows considering the location and weather dependencies of non-CO2 species, as well as forecast-related uncertainty in conversion to equivalent CO2 emissions. Simulation results demonstrate the effectiveness of the proposed in optimizing aircraft trajectories, yielding mitigation in both climate impact and operating cost.

2. Conflict Resolution of Climate Optimal Trajectories using Reinforcement Learning

The second conference paper, led by Fateme Baneshi, explored a strategic deconfliction framework based on multi-agent reinforcement learning to address aircraft conflict resolution in high-density traffic scenarios. The proposed strategy leverages a deep deterministic policy gradient algorithm to provide speed advisory to operating aircraft with the aim of minimizing the number of conflicts. To address the non-stationary environment challenge, a centralized learning, decentralized execution scheme is implemented. In the training process, the agents make use of information from other aircraft, but this information is not utilized during the testing phase. The effectiveness of the proposed approach is validated by analyzing various sets of climate-optimal trajectories, which pose critical safety issues due to the uneven distribution of flights within the airspace. The results demonstrate that the proposed framework effectively resolves a high number of conflicts that arise as a result of adopting climatically-optimal trajectories.

An effective showcasing platform

The Aerospace Europe Conference in Lausanne proved to be a useful platform for researchers and industry experts to showcase their latest advancements and innovations in the aerospace field. The contributions of the RefMap team from Carlos III University shed light on the potential of climate-optimal flight planning and conflict resolution strategies to take significant steps towards a more sustainable and environmentally responsible aviation industry.

bottom of page