Page 26 - ATC Special Bulletin Series - Training & Simulation 2022-01
P. 26

2. Many small projects acting independently; and
3. Alackofconsistencybetweenthedifferentprojectsinhow project objectives were set and how results were captured.
The E-OCVM is focused on stages V1-V3 of the concept lifecycle model (i.e. the scope, feasibility and pre-industrial development and integration phases). E-OCVM was last updated in 2010 and since that time, we have seen the explosion of advanced computational techniques such as Artificial Intelligence (AI) and Big Data. Our ATM research community are developing ATM concepts that rethink the distribution and responsibility of tasks between the human and the machine. We also see the increased alignment and collaboration between sovereign nations enabling new concepts and benefits that operate on a continental scale.
While we believe that the fundamentals of the E-OCVM remain robust, it must be acknowledged that the context under which the methodology was established is different to the aviation world we see forming today. Using today’s validation methodology for future ATM concepts and technologies is therefore likely to miss issues within a concept ‘lurking below the surface’ and allow risk to be carried through to the implementation phase. Without action we believe that the risk mitigation and assurance offered by validation over the last 15+ years will diminish, and implementers and regulators will not benefit from the validation tools and processes that have successfully helped deliver innovative concepts into deployment.
Let’s consider three different emerging areas in which the application of the current E-OCVM toolset may not provide comprehensive support to effective validation:
1. Application of AI for Automation of ATC :
The application of AI to help (or fully) automate ATC tasks is one of a growing set of R&D solutions that are primarily technology- focussed and if successful will introduce considerable change to human roles in ATC. The E-OCVM has historically been orientated towards operational concepts that optimise procedures or provide computer aids to users in their existing roles, rather than reframing them. Going forward equal consideration of technological and operational concepts are needed within the methodology for it to satisfy the needs of technologies of ever- expanding complexity.
For learning technologies such as AI, an understanding is needed that the benefits provided by the solution will be variable once deployed, based of service duration and quality of data inputs. Up to this time, the focus for R&D validation within ATM has been on V1 to V3 concept maturity levels, and

   24   25   26   27   28