Before starting:
  * Back up /etc/rts2/ir.model
  * Make sure the script for target 4 is appropriate, at least two exposures to allow astrometry to complete before the next pointing.
  * Clear the existing model with 'rts2-opentpl-errors -I <ip> -N <port> --reset-model'
To create the new model:
  - rts2-mon/EXEC:
    * Verify auto_loop and default_auto_loop are true
    * Execute command 'now 4'
  - shell:
    * Track the number of points with 'rts2-opentpl-errors -I <ip> -N <port> --sample'
    * After $ENOUGH points, do 'rts2-opentpl-errors -c' to calculate (and load) the new model.
  - rts2-mon/<tel>:
    * Execute command 'save_model' (to write new /etc/rts2/ir.model

Astelco suggest starting with a 7 point model.  Martin Jelinek suggests finishing with 50 points.