On the second day of the first HAAWAII stakeholder workshop (29th June 2021) the Human Performance (HP) working group took place. The session was facilitated jointly by HP specialists from both Austro Control and NATS. The aim of the working group was to share the insights gained so far in speech recognition as a valuable HP metric and more importantly to get the participants’ feedback and ideas on the way forward.
The first half of the session was used to present participants with HP concepts as well as opportunities and challenges of using speech recognition output as a proxy for overall human performance. The big question that we put in front of participants was “How would you quantify workload based on speech recognition data?”.
The second half of the session was used to discuss this question as well as participants’ views on the presented material overall.
The main findings were:
- Acknowledgement that a joint approach between quantitative and qualitative data gathering and analysis is needed to produce accurate insights into HP
- Recognition that extensive validation is needed with both traditional and new tools
- Assurance will be required before speech recognition data can be used for decision making by users
- Work on target audience is needed – data needs to be presented to supervisors and network planners and developed with a user centered design philosophy in mind
- Speech recognition data should be interwoven with other decision data such as weather or complexity
- Desire to use speech recognition data for off-line design like airspace or technology developments and change management
- Preference for simple visual displays of data rather than numbers or graphs
- Integration obstacles remain; especially convincing ANSPs or users to have unified data strategies
- Need to train supervisors and other users on how to trust the use of big data for decision making
In summary, the working group was well received and confirmed as well as broadened the HAAWAII team’s understanding of how to progress the use of speech recognition data as another puzzle piece in measuring overall human performance.