MIT Research on Randomizing Tasks

I worked on a project that looked at how job tasks could be randomized in industrial processes.  This research effort at MIT, fairly technical, looks at a more arcane area of massive multicore processors, but still touches on how you can use context knowledge to randomly direct attention.   Could this be used for human task attention?