Periodic Reporting for period 2 - ENVISION (Enhanced Situational Awareness through Video Integration with ADS-B Surveillance Infrastructure on Airports)
Período documentado: 2019-01-01 hasta 2019-12-31
The ENVISION project proposes an innovative low-cost solution for real-time tracking of aircraft and vehicles at airports. Real-time tracking has become an essential requirement for airports to allow enhanced safety, increased controller awareness and optimized airport operations. Simply speaking, the system provides a ground radar image without radar.
When compared to ENVISION, existing tracking solutions rely on costly and complex infrastructure (Radar, multiple antennae) which not cost effective for small and regional airports.
The proposed ENVISION solution relies on technologies such as video cameras, aircraft transponder positions (ADS-B) and LiDAR sensors (as used in autonomous cars for detecting obstacles and discovering the surrounding environment). In others terms it relies on multiple surveillance sensors to complement and supplementing each other.
ENVISION feeds the controller display, with the most reliable surveillance information from the multiple sources available at any time, including provision for sensor failure.
As such, the project aims to address the following detailed objectives:
- Assess operational impact of using CCTV camera and LIDAR as surveillance sensors providing identification and positioning data for airport surveillance application and airport collaboration tools.
- Define initial safety and performance requirements for use of CCTV and LIDAR as surveillance sensors supplementing ADS-B and required surveillance standard updates.
- Provide a detailed design of the surveillance infrastructure, including technical requirements for CCTV camera, LIDAR sensor, and specification of innovative image processing and data fusion algorithms and interfaces to the ATC and Airport systems.
- Implement surveillance and collaboration interface prototypes integrating the ADS-B, CCTV and LIDAR surveillance sensors.
- Evaluate operational benefits through simulation, demonstration, and live trials in a stepwise approach by validating the implementation in a small airfield and a regional airport.
- Analyze the cost-benefits ratio for such solution through a paper study fed by interviews with airport stakeholders.
When compared to ENVISION, existing tracking solutions rely on costly and complex infrastructure (Radar, multiple antennae) which not cost effective for small and regional airports.
The proposed ENVISION solution relies on technologies such as video cameras, aircraft transponder positions (ADS-B) and LiDAR sensors (as used in autonomous cars for detecting obstacles and discovering the surrounding environment). In others terms it relies on multiple surveillance sensors to complement and supplementing each other.
ENVISION feeds the controller display, with the most reliable surveillance information from the multiple sources available at any time, including provision for sensor failure.
As such, the project aims to address the following detailed objectives:
- Assess operational impact of using CCTV camera and LIDAR as surveillance sensors providing identification and positioning data for airport surveillance application and airport collaboration tools.
- Define initial safety and performance requirements for use of CCTV and LIDAR as surveillance sensors supplementing ADS-B and required surveillance standard updates.
- Provide a detailed design of the surveillance infrastructure, including technical requirements for CCTV camera, LIDAR sensor, and specification of innovative image processing and data fusion algorithms and interfaces to the ATC and Airport systems.
- Implement surveillance and collaboration interface prototypes integrating the ADS-B, CCTV and LIDAR surveillance sensors.
- Evaluate operational benefits through simulation, demonstration, and live trials in a stepwise approach by validating the implementation in a small airfield and a regional airport.
- Analyze the cost-benefits ratio for such solution through a paper study fed by interviews with airport stakeholders.
Different dissemination activities were performed : project website, presentation in several international events (World ATM Congress 2018 & 2019, Dubai Airport Show 2018 , ProAvia Conf. 2018 , Assises du Développement durable in Toulouse, SID 2018/2019), discussions with several stakeholders and demonstration workshop at Carcassonne airport.
The Operational work package identified the state of the art (sensing technologies & applicable standards). Operational and an initial set of safety requirements together with potential impact on standardization materials was identified. The regional airport A-CDM concept has been detailed and uses cases were identified.
As part of the Technical work package, a complete surveillance solution was designed. The ENVISION system integrated sensors (cameras, LIDAR and ADS-B) where the surveillance data, consolidated by the fusion module, fed the A-SMGCS and the A-CDM user interfaces. Consolidated surveillance data also fed the A-CDM user interface.
Several LiDAR detection and Video Processing algorithms were assessed.
The DataFusion module centralized the surveillance data from the different sensors and computed the most accurate position possible for every detected target (aircraft, vehicle or person) in the covered area.
The target position reports, and the flight milestones events (AOBT, ATOT, ALDT, …) fed the two applications prototypes - an A-SGMCS and an A-CDM GUI, designed for the project.
ENVISION A-CDM is intended to support the EUROCONTROL Advanced ATC Tower CDM concept for regional airports. This means that the Regional Airport will be able to receive accurate arrival times from and send accurate departure times to the EUROCONTROL Network Manager (NM) via their B2B Web Services. ENVISION A-CDM is a browser-based web application that is accessible through any device.
During the first year, all the trials, validation and demonstration activities (WP4) were held in Muret airfield (ICAO: LFBR). During the second year, the system has been installed at Carcassonne airport (ICAO: LFMK) and assessed with commercial flight.
For the first assement in Muret, different scenarios, involving cooperative and non-cooperative targets have been defined and several recording sessions have been done to feed the WP3 developments.
The final validation deployment in Carcassonne permitted to evaluate the performance of the system for commercial aircraft.
A CBA study (WP5) provided an overall understanding of the capital and operating costs involved for deploying an ENVISION-based A-SMGCS and A-CDM solution to airports in Europe and evaluates the benefits that the solution provides in terms of safety and airport capacity. When possible, this CBA identifies the categories of airports (by size, by experienced weather conditions, by airport layout complexity, etc.) that should most benefit the considered solution.
The objective is to demonstrate a CBA model for a set of airports in Europe so this model can be used by European airports as a reference when developing their own business case.
The Operational work package identified the state of the art (sensing technologies & applicable standards). Operational and an initial set of safety requirements together with potential impact on standardization materials was identified. The regional airport A-CDM concept has been detailed and uses cases were identified.
As part of the Technical work package, a complete surveillance solution was designed. The ENVISION system integrated sensors (cameras, LIDAR and ADS-B) where the surveillance data, consolidated by the fusion module, fed the A-SMGCS and the A-CDM user interfaces. Consolidated surveillance data also fed the A-CDM user interface.
Several LiDAR detection and Video Processing algorithms were assessed.
The DataFusion module centralized the surveillance data from the different sensors and computed the most accurate position possible for every detected target (aircraft, vehicle or person) in the covered area.
The target position reports, and the flight milestones events (AOBT, ATOT, ALDT, …) fed the two applications prototypes - an A-SGMCS and an A-CDM GUI, designed for the project.
ENVISION A-CDM is intended to support the EUROCONTROL Advanced ATC Tower CDM concept for regional airports. This means that the Regional Airport will be able to receive accurate arrival times from and send accurate departure times to the EUROCONTROL Network Manager (NM) via their B2B Web Services. ENVISION A-CDM is a browser-based web application that is accessible through any device.
During the first year, all the trials, validation and demonstration activities (WP4) were held in Muret airfield (ICAO: LFBR). During the second year, the system has been installed at Carcassonne airport (ICAO: LFMK) and assessed with commercial flight.
For the first assement in Muret, different scenarios, involving cooperative and non-cooperative targets have been defined and several recording sessions have been done to feed the WP3 developments.
The final validation deployment in Carcassonne permitted to evaluate the performance of the system for commercial aircraft.
A CBA study (WP5) provided an overall understanding of the capital and operating costs involved for deploying an ENVISION-based A-SMGCS and A-CDM solution to airports in Europe and evaluates the benefits that the solution provides in terms of safety and airport capacity. When possible, this CBA identifies the categories of airports (by size, by experienced weather conditions, by airport layout complexity, etc.) that should most benefit the considered solution.
The objective is to demonstrate a CBA model for a set of airports in Europe so this model can be used by European airports as a reference when developing their own business case.
Technical performance:
The system deployed at Carcassonne did not allow to reach A-SMGCS performance requirements because of the limited number of sensors (3 cameras & 1 ADS-B receiver) but still provided very promising results.
A wide-angle camera (>100 degrees) allows tracking of a commercial aircraft within a range of 350 m, with a good probability of detection (>90% every second). Limitations observed could alleviated with denser sensors deployment and further training of the deep learning model.
With a short detection range (100-200m) but a wide angle (360 degrees), LiDAR could efficiently complement deployments as it enhances the accuracy.
Video-based object recognition can track all kind of objects including vehicles, persons and obstacles. This classification capability could be also extended for airline detection, aircraft model, etc.
Cost-benefits analysis:
The CBA only covers a limited portion of all operational benefits, and are thus clearly underestimated, the study concludes that an ENVISION-based A-SMGCS system is :
- a valid alternative for airports experiencing between 100 000 to 150 000 movement per years, offering a reasonable payback period (<5 years)
- a possible alternative for airport experiencing between 50 000 to 100 000 movement per years, although inducing a significant payback period (>5 years)
- unsuitable for small-size airports, i.e. for airports experiencing less than 50 000 movements per years
The CBA also underlines that alternative sensors to LiDAR should be investigated to improve the solution’s benefit/cost ratio:
- the LiDAR sensor’s unit cost is high (~10 k€) & a massive deployment is required, it also suffers limitations: sensitivity to fog, heavy rain and snow
Operational:
The system was generally able to track mobiles and people using cameras + ADS-B data. Whilst the Surveillance performance (accuracy and lack of identification) did not meet the A-SMGCS requirements, it was encouraging to see mobiles being tracked without the installation of traditional A-SMGCS.
A-CDM data for aircraft with flight plans was generally accurate and alerts were generated, especially for late arrivals. Ghost DPI messages were generated.
The system deployed at Carcassonne did not allow to reach A-SMGCS performance requirements because of the limited number of sensors (3 cameras & 1 ADS-B receiver) but still provided very promising results.
A wide-angle camera (>100 degrees) allows tracking of a commercial aircraft within a range of 350 m, with a good probability of detection (>90% every second). Limitations observed could alleviated with denser sensors deployment and further training of the deep learning model.
With a short detection range (100-200m) but a wide angle (360 degrees), LiDAR could efficiently complement deployments as it enhances the accuracy.
Video-based object recognition can track all kind of objects including vehicles, persons and obstacles. This classification capability could be also extended for airline detection, aircraft model, etc.
Cost-benefits analysis:
The CBA only covers a limited portion of all operational benefits, and are thus clearly underestimated, the study concludes that an ENVISION-based A-SMGCS system is :
- a valid alternative for airports experiencing between 100 000 to 150 000 movement per years, offering a reasonable payback period (<5 years)
- a possible alternative for airport experiencing between 50 000 to 100 000 movement per years, although inducing a significant payback period (>5 years)
- unsuitable for small-size airports, i.e. for airports experiencing less than 50 000 movements per years
The CBA also underlines that alternative sensors to LiDAR should be investigated to improve the solution’s benefit/cost ratio:
- the LiDAR sensor’s unit cost is high (~10 k€) & a massive deployment is required, it also suffers limitations: sensitivity to fog, heavy rain and snow
Operational:
The system was generally able to track mobiles and people using cameras + ADS-B data. Whilst the Surveillance performance (accuracy and lack of identification) did not meet the A-SMGCS requirements, it was encouraging to see mobiles being tracked without the installation of traditional A-SMGCS.
A-CDM data for aircraft with flight plans was generally accurate and alerts were generated, especially for late arrivals. Ghost DPI messages were generated.