Trajectory Prediction in Air Traffic Management

Trajectory prediction in air traffic management encompasses the forecasting of aircraft paths in four dimensions—latitude, longitude, altitude and time—over various flight phases. Accurate prediction underpins crucial operations such as conflict detection, sequencing and airspace capacity planning. Traditional methods have relied on physics-based kinematic models and state estimation, but recent demands for higher airspace efficiency and safety have driven the adoption of data-driven and hybrid approaches. These incorporate machine learning techniques to model complex dependencies arising from aircraft performance, environmental uncertainties and inter-aircraft interactions. The global proliferation of high-resolution surveillance data, including ADS-B feeds, has enabled the development of models that balance long-term trend estimation with responsiveness to tactical manoeuvres. Real-time integration of predictive outputs into decision-support tools promises enhanced operational resilience and reduced controller workload.

Research from Nature Portfolio

Recent studies have introduced a wavelet-based time-frequency analysis framework integrated into an encoder–decoder neural network to capture both global flight trends and local manoeuvre details, demonstrating marked accuracy gains during climb and descent phases. Another approach combines long short-term memory networks with attention layers to weigh temporal features selectively, thereby enhancing 4D trajectory forecasts under waypoint sparsity and operational uncertainties. These innovations illustrate how the fusion of frequency-domain methods and adaptive weighting mechanisms can yield robust, high-precision predictions in dynamic airspace environments.

Trajectory Prediction in Air Traffic Management publication trend

The graph below shows the total number of articles in trajectory prediction in air traffic management across all publications each year (not limited to Nature Index journals).

Technical terms

4D trajectory: A representation of an aircraft’s path including longitude, latitude, altitude and time.

Wavelet transform: A time-frequency analysis technique that decomposes signals into components at various scales to reveal local and global patterns.

Attention mechanism: A neural network module that assigns dynamic weights to input features, emphasising the most relevant information.

Encoder–decoder architecture: A sequence-to-sequence neural framework where an encoder processes input data into a latent representation and a decoder generates the predicted sequence.

ADS-B (Automatic Dependent Surveillance–Broadcast): A surveillance technology whereby aircraft periodically broadcast their position and state information.

References

1. Flight trajectory prediction enabled by time-frequency wavelet transform. Nature Communications (2023).
2. Attention-LSTM based prediction model for aircraft 4-D trajectory. Scientific Reports (2022).
3. Multi-aircraft attention-based model for perceptive arrival transit time prediction. Advanced Engineering Informatics (2025).
4. 4D flight trajectory prediction using a hybrid Deep Learning prediction method based on ADS-B technology: A case study of Hartsfield–Jackson Atlanta International Airport (ATL). Transportation Research Part C Emerging Technologies (2022).
5. Aircraft Trajectory Prediction With Enriched Intent Using Encoder-Decoder Architecture. IEEE Access (2022).