March 19, 2003
Attendees: Trish Ververs, Jim Sampson, Ben Darling, Jason Pratt, Jesse Schell, Allison Styer
; Gauge validation for vigilance at CMU
Roy Stripling, Don DuRousseau, and another HB employee traveled to CMU to collect data on the gauges (XLI, engagement, IBI, and arousal meter, P300?). The data will be provided by CMU shortly. They ran into problems with the code. Kyle was able to contact Roger and resolve some of the issues, however there appeared to be a memory leak that Roger Carff couldn’t find. They did collect the gauge data but there was little to no raw EEG data collected.
; Use of MoCap lab for non-VE experiments
There was concern that the MoCap lab was used for a non-immersive VE experiment. All the equipment was there and the BioSemi system was there. There was little time to coordinate after the CMU spring break. In the future, if a desktop experiment needs to be run they will move the equipment out of the lab. Honeywell doesn’t anticipate running another desktop evaluation at CMU. The next one should be in the MoCap environment. Desktop evals should be run at IHMC. Trish will check with Cortech (Lloyd Smith) about arrival of BioSemi equipment. On Feb 24, Lloyd said 4 weeks which is next week.
Update: Lloyd called BioSemi and in addition to the delay in electrodes there is now a slow down on the delivery of motherboards. Lloyd swapped us with another order of someone who is less anxious so we are bumped up in the queue by 2 weeks, but they still anticipate a 2-week slip – delivery April 15. There are a couple of things that BioSemi can do to speed up the process and Lloyd strongly encouraged them to do so, so the delivery could be sooner.
; Change in MoCap vigilance task
There is a notable difference between the desktop vigilance task and the MoCap vigilance task. The events are only available in the desktop for 2 seconds where the MoCap gives an unlimited amount of time to detect the targets. In pilot studies at CMU they are failing to find the vigilance decrement. It was decided to mimic the desktop and only have the targets available for a limited period of time. Choices are popping up in a window or only walking by a single window. They will try both (with a slight preference for walking across the window). They’ll need to make sure they can get the vigilance decrement first and then work on having the subject target (shoot) the foe. It was suggested that they test using a mouse response first. They will also try to have at least 30 events (friends in windows) per minute. They will be testing a number of iterations (4) on this task with naïve subjects this Sunday.
; Change in mitigation strategy
Honeywell tested the first proposed mitigation of the vigilance decrement with 2 subjects and performance was NOT improved. It actually decreased. Ben agreed that the secondary task was competing for resources (created a divided attention task) and that this type of mitigation (i.e., report if there is a change in the display) is better for longer period vigilance tasks where the target is present for a long period of time. We decided to go back and try increasing the event rate of a secondary task. So instead of trying to remember the past configuration, all the subject has to do is look for another type of target. For instance, look for Ts and upright Ls. The equivalent mitigation in the MoCap would be to search for foes and friends with some kind of feature (holding a gun, black mask, etc).
; April 13 at CMU
The meeting at CMU is confirmed for April 13. At this time, Dylan will not be able to attend but Amy and the Admiral will. Henry Girolamo and Jim Sampson will also be present.