Enhancing Continuous Descent Operations Efficiency through AI

Fuel efficiency and sustainability are critical concerns in the aviation industry, requiring innovative approaches to address them. With the increasing demand for air travel, managing this growth efficiently becomes crucial to minimizing the environmental footprint. Optimizing flight paths, especially through the implementation of Continuous Descent Operations (CDO), offers a significant opportunity to reduce fuel consumption and carbon emissions. By leveraging artificial intelligence (AI), we can further enhance these outcomes predicting optimal routes to manage air traffic more effectively.

Strengthening Air Traffic Management Through Collaboration

A strategic partnership between the Air Traffic Management Research Institute (ATMRI) and AIR Lab is helping to close the gap between theoretical research and real-world ATM applications.

ATMRI, founded in 2013 by the Civil Aviation Authority of Singapore (CAAS) and Nanyang Technological University (NTU), focuses on regional air traffic innovation. Its research spans digital towers, human–AI collaboration for controllers, advanced analytics, urban aerial mobility, and emerging technologies such as human-AI systems and large language models.

AIR Lab, created in 2019 as a joint venture between CAAS and Thales, develops open ATM system architectures. It brings together stakeholders across the ATM value chain — including controllers, engineers, and technology partners — to advance green aviation initiatives using open technologies and artificial intelligence.

AI & Machine Learning Enhancing CDO Operations

Under ICAO standards, CDO allows aircraft to follow a descent profile optimized to the aircraft’s operational capabilities, using low engine thrust settings to reduce fuel burn and emissions.

Determining the optimal Top of Descent (TOD) for CDO is complex. It must account for aircraft type, real-time and forecast weather, terminal maneuvering area (TMA) constraints, and evolving traffic patterns. Key input factors include:

• Aircraft performance characteristics
• Real-time and forecast weather
• Airspace structure and constraints
• Dynamic traffic patterns

Machine learning models can ingest historical and real-time data from successful and unsuccessful CDO operations to refine TOD predictions. By integrating data from Air Traffic Management Systems, these models continuously learn and improve, enabling more precise, fuel-efficient descent profiles.

As ATMRI researchers highlight, AI and ML can significantly enhance TOD calculations for CDO, resulting in more accurate and environmentally efficient approaches.

Collaborative Value Outcomes

The collaboration between AIR Lab and ATMRI demonstrates how academic research can be translated into operational ATM solutions. By combining research insights with industry expertise, the partnership is improving air navigation services and setting new benchmarks for aviation efficiency.

Ultimately, ATMRI and AIR Lab show that sustainability and operational excellence are mutually reinforcing goals. With AI as an enabling technology, the aviation sector can pursue both simultaneously, achieving greener operations without compromising performance.