Assigning parallel jobs to less busy cores on a shared windows server

Question

In parallel processing in R (e.g., using future) under Windows, is it possible to force R to prioritise the use of less-busy cores on a shared device? I am using a shared Windows server (> 200 cores) with many other colleagues. some of the parallel jobs were killed, probably because the same cores used by R are already in use by other colleagues. Is it possible to dynamically avoid using cores with e.g. > 80% utilization (see attached snapshot)?