|
|
|
|
Background |
|
|
|
|
|
|
|
|
|
The Air Traffic Management (ATM) systems within the European Union and US are complex, dynamic, information-driven systems based on a human-centred automation paradigm. The human operators of these ATM systems make decisions based on information produced, in large part, by machines. Trajectory Prediction (TP) is a critical component of this information.
Current ATM modernisation programs are deploying automation systems and Decision Support Tools (DSTs), each with their own uniquely designed and integrated TP capabilities. Given differences in design, even with the same data, each TP module may yield different trajectory predictions with different levels of computational effort. Independent efforts to date have been advantageous in that they have enabled fast parallel progress. For current operational applications, the ATM community has enjoyed success in deployment and operational use of the current generation of DST automation capabilities across many Air Traffic Service Providers (ATSPs).
Going forward, TP performance will be even more critical to meeting the requirements of future ATM DST clients. The deployment of new DST capabilities could increase the number of TP modules that are deployed. For future operational applications, this situation may present ATM stakeholders with several potential concerns that must be assessed. First is the interoperability of the overlapping TP information (i.e., do operationally significant differences in results exist?). The concern here is that any one controller could potentially be provided with complementary and/or overlapping information generated from multiple TP modules. Second, although current TP technology appears sufficient to support the first generation of Basic DST applications, evidence suggests[1] that the performance of current TP capabilities may not be sufficient to support more advanced automation applications. Third is the cost of research, development, deployment, and maintenance of multiple supporting TP modules that are similar if not overlapping. Fourth is the issue of developing cost-effective accuracy improvements and requirements for data dissemination (e.g. adaptation data in suitable formats, aircraft performance data, etc.) to provide a desired level of system performance.
Very little objective data exists within the community’s literature to help stakeholders answer these critical and potentially costly questions. Individual automation projects must bear the burden to develop and validate their own TP capabilities. If the practice of independent efforts continue, progress may slow and become more costly because of duplication and resource limitations. This paper proposes a community effort to collaborate on the development and validation of common TP-supporting technologies and information. It is felt that this approach will reduce the cost of DST R&D, improve quality and interoperability, while freeing up individual stakeholders to focus their efforts, resources, and expertise on the unique features of their DST technologies.
- [1] ”Operational Requirements for Trajectory Prediction for EATCHIP Phase III”, Eurocontrol Doc. OPR.ET1.ST03.1000-ORD-02-00
|
|
|
| |
|
 |
|
Last validation: 23/12/2004
|
|
|
|
|
|
|
