This work is focused on norm operationalization in aviation domain. The investigated paradigm can be described as: from legal norms to technical rules in the artifact. Normative requirements (norms) for the aircraft trajectories are extracted from the flight rules and airport procedures, and operationalized in a decision support system (DSS). The decision support is based on evaluation of risk to violate the normative requirement. The following risks are modeled: trajectories' conformance with the flight rules, safe distance between aircraft, wake turbulence separation and avoidance of volcanic ash. The DSS is for the air traffic controller (not pilot) and must respond in real time. It provides surveillance, evaluates and recommends, whereas the human controller takes a decision.
This research is focused on the operationalization of normative rules in aviation domain (air traffic control, ATC). A proposed paradigm can be called “from legal norms to technical rules in the artifact”. Normative requirements are extracted from the flight rules, maps and approach/departure procedure charts. An example of a normative rule is “Keep 3 degrees descent angle while landing and hold restrictions of the altitude and geography depicted in the aerodrome chart”.
Normative rules are modeled in order to provide decision support in terms of norm
violation risk. A decision support system (DSS) provides surveillance, evaluates and
recommends, whereas the human controller takes a decision. The final decision is done
by human controller. This approach accords with SESAR (Single European Sky Air
Traffic Management Research) target concept, which states that humans should
constitute the core of the future air traffic management (ATM) operations [
The decision support is based on lidar (laser radar – LIght Detection And Ranging) and radar data fusion. It relies on the assumption that the precise aircraft position data (with error margin of meters, not hundreds of meters) from the lidar will facilitate detection of risks that are not possible to detect using only radar data.
The research goal is to develop a conception for operationalization of the aircraft approach/departure norms in a decision support system, taking into consideration the use of lidar for aircraft tracking. The goal is broken down into these tasks: (1) modeling norm violation risk in the airport traffic zone (ATZ), (2) modeling radar and lidar data fusion, and (3) development of a prototype decision support system.
The norm operationalization is investigated in the context of the ATM paradigm developed in the EU FP6 SKY-Scanner project1: expanding surveillance and the ATC control to the approach/departure phases by using radar and lidar data fusion and decision support in terms of norm violation risk. Only norms that can be checked using the lidar-radar fused data (position and speed) are examined and included in the operationalization conception.
The project was aimed at developing a laser system to detect and track aircraft up
to at least 6 nautical miles (NM) from the aerodrome traffic zone (ATZ) barycenter [
Several important assumptions stem from the SKY-Scanner project and thus form the boundaries of the current research:
1) ATC activities require a real-time response from the DSS. A study of
timecritical decision support models provided in [
2) The emphasis is on informing the controller, who then makes a final decision on
the actions. This accords with the results of studies of human-automation interaction:
high levels of automation are not advisable in systems dealing with dynamic
environments with many external and changing constraints [
Normative rules for aircraft approach/departure from International Civil Aviation
Organization (ICAO) flight rule documents [
Current aviation-related decision support systems do not model norms
comprehensively, but there is some research in that direction. One type of decision
support – Conflict Detection and Resolution (CD&R) systems. The structure of the
1 “Development of an Innovative LIDAR Technology for New Generation ATM
Paradigms” (SKY-Scanner), 2007-2010, http://www.sky-scanner.it/
CD&R process [
2D visualizations in the ATM domain are no longer sufficient, and the modern 3D
visualizations have drawbacks [
The approach/departure decision support focuses on detecting violations of the flight rules for the aircraft. We conceptualize each norm as a triplet of a norm factor, norm pattern, and the expected value. Norm factor represents a quantitative trajectory attribute of one or several aircraft. Only factors that can be computed from the DSS input data are considered in this conception. Expected value, vN, is the value defined in the text of the normative requirement. Norm pattern (‘≤vN’, ‘≥vN’, or ‘=vN’) explicates how to interpret the expected value. For example, norm pattern ‘≥vN’ means the actual value of the factor should be greater than the expected value.
Patterns ‘≤vN’ and ‘≥vN’ constitute limit-based norms, and pattern ‘=vN’ –
deviation-based norms. Example of the limit-based norm (Figure 2): “height minimum
is 3900 ft. at 6 nautical miles from distance measurement equipment (DME)”. An
example of the deviation-based norm could be the track (the direction that the aircraft
should follow), which is expressed in degrees from North, e.g. 236° [
The defined operationalization structure translates each norm into a risk definition in the DSS. The use of discrete risk levels abstracts from unnecessary details. In the DSS risk levels are defined based on the likelihood of violating the norm. A separate risk definition is formulated for each normative requirement. An individual risk evaluation maps the observed factor value to a discrete scale of risk levels.
The L-level risk concept is characterized by five elements (Figure 3): (1) risk factor (e.g. ‘altitude’ or ‘indicated airspeed’); (2) risk type (‘limit’ or ‘deviation’); (3) the norm pattern (‘≥vN’, ‘≤vN’, ‘=vN’); (4) expected value of the factor; (5) a set of thresholds for risk levels. If the risk type is ‘limit’, a set of thresholds consists of L-1 constants, defined in the terms of factor measurement units. If risk type is ‘deviation’, a set of thresholds consists of L-1 pairs of constants, defining allowable deviation levels.
For convenience of visual representation of the risk definition, a piece-wise linear
risk-magnitude function is used, which maps the observed factor value to a number
from the interval [
As an example we consider the norm “altitude 3900 ft at 6 DME” (Figure 2). In deviation-based risk evaluation we consider not the expected value itself, but the expected deviation (dN = 0). There are 4 risk levels and 3 pairs of thresholds. The corresponding risk definition is: (1) factor: 'altitude'; (2) type: 'deviation'; (3) pattern: ‘=vN’; (4) expected value: 3900 ft at 6 DME (deviation 0); (5) thresholds: dn0 = -0.5, dp0 = 2, dn1 = -1, dp1 = 3.5, dn2 = -1.5, dp2 = 5; (see Figure 4) The threshold values in this example are chosen only for demonstration purposes.
Each risk is represented in a separate indicator on the DSS control panel (Figure 5). The risk level is shown with color and the number of colored slots on the indicator.
The DSS prototype embodies the norm operationalization conception proposed in the previous chapter. It illustrates the modeling of several norms for the approaching aircraft, and provides a real-time simulation of the suggested decision support scenario.
Advanced ATC visualization ideas are adapted to context of approach/departure
decision support. The DSS prototype provides a laboratory implementation, which
advances technology readiness level (TRL [
2D-in-3D prototype uses generalized terrain model and embeds 2D semitransparent projection walls (Figure 6). Aircraft are represented with spheres. The ought-to-be trajectory is projected on the wall (white line), as well as the aircraft position (black dot). If the dot is not on the line, there is a path violation.
The pure-3D prototype uses photographic terrain (high resolution satellite images) and represent the aircraft as full 3D models (Figure 7). 2D rings enclose the ought-tobe trajectory (plus some allowable deviation). The violation is detected when the aircraft indicator is outside the rings. This approach is less strict that the 2D-in-3D.
Human operator needs are satisfied in the following way: 3D display improves situation awareness as the airport environment is depicted with essential terrain obstacles; 2D elements (walls, ring) that relate the aircraft position to the airport procedures reduce cognitive workload.
The following conclusions are drawn: 1. The proposed norm operationalization conception enables to represent a subset of aircraft approach/departure normative rules in a decision support system for the air traffic controller. The referred subset is defined as the norms concerning aircraft trajectories, or simply, geometrical norms. 2. The prototype decision support system provides an integrated solution to facilitating the controller: risk indicators automate detection of the possible norm violations, 2D-in-3D visualizations help comprehend conformance to the approach/departure procedure. The controller’s work improvement is not quantifiable because the research addresses the needs that will only become relevant in the future. 3. Analysis of the prototype development process demonstrates that the following steps are needed to operationalize a norm: (a) setting up risk levels and colors, (b) creating risk definitions (consisting of norm factor, expected value, type, pattern and a set of thresholds), and (c) setting up risk indicators. The process cannot be fully automated, as each norm factor is unique, and analysis has to be performed for each new kind of norm to be operationalized.