Results from the First Test Event of DecisionLab
Just an update on DecisionLab. Well! we ran the course at the College of Policing last week and it was great. The course was the MAGIC course with representatives from Police, Fire, Health, Local Authorities, Utilities, Government and the Military.
The approach is to replicate a multi room Hydra exercise in the same room, with the participants sitting round tables representing the syndicate rooms, all wearing headsets, all connected to the DecisionLab system.
I can report that this mountain has been climbed. The event was very well run by the College of Policing team and the students, through DecisionLab, were deeply immersed into a Hydra major incident strategic management. The data was captured effectively from the system and I have already started to pass it through my AIQA analysis tools.
Lots to do, now I have some real data but here is a sniff of the achievable (this is a first pass to test of some of the analysis approaches I am developing . The table discussions analysed against decision logs generated some very insightful data. I am looking at a number of components:
1. Participants and Roles
2. Key Themes and Concerns
3. Group Dynamics
4. Emotional Tone
5. Decision-Making Process
6. Gaps and Uncertainties
Snippet of findings, so far, lots to do here!
Decision Making Process - AIQA response - Snippet
The incident's evolving nature, including unconfirmed casualty numbers and uncertain structural safety, causes anxiety and caution. The Commander shows some hesitancy when discussing the risk assessment and is cautious in committing to strategic language.
This is evidenced by moments where the Commander questions the wording of strategic aims, indicating doubt or lack of confidence in certain decisions.
Recommended by LinkedIn
Comparison with the written logs:
The decision logs do not reflect this sense of doubt or apprehension. The tone in the logs is entirely procedural, presenting decisions as if they were made with full confidence and clarity.
There is little of the emotional tension that occurs when making decisions in uncertain and rapidly changing situations.
Next Steps
I have proved the concept works, now I have the captured data from all the tables, the strategic meetings, the decision logs and the press conference. My focus here is now building the analysis tools, the capture mountain has been climbed!
International Oil Trader retired speaking fluent Spanish, German and English mother tongue.
1moInteresting event
Global Head of CTI
1moIncredible evolution…. I wouldn’t expect anything else. Congratulations Jonathan Crego and all concerned.
Hi Jonathan. Just saw this and had a look at the info on the link. It looks like you are recording/transcribing the live conversations and AI-ing (is this really a verb!!!) them for comparison with the written record. Brilliant piece of work. I am currently of the opinion that we should not record or transcribe live crisis management settings as these need to be psychologically safe spaces were the unthinkable needs to be voiced and perhaps discussed. Recording and transcription can leak, or be correctly released, but subsequently misused - particularly by those who slice a quote out of context. Do you agree? I'm sort of hoping you don't 😊 All my best as ever. Sean
Research Associate (RA) at the University of Liverpool & Honorary RA at Cardiff University / Prifysgol Caerdydd
1moMany congratulations Jonathan! The initial data analysis is fascinating 👍
--
1moLove this, Jonathan