Hello Jesus and Joel,
We removed the hands-on mode this year. We had one experimental setup
in the lab which the students could "touch and feel", but otherwise
we asked them all to use the iLab setup. One reason for this is the
small differences (mainly friction) in the different experimental setups.
Hence, the students may get it working on the lab setup, but then have
to make modifications on the iLab setup.
I will be marking the practical reports in the coming weeks, and I will
email Joel the statistics of which groups were successful, etc.
The lessons from this year's run from my point of view are listed below:
1. The iLab experiment was online for 5 weeks, yet we had virtually no
usage in the first 4 then a massive run during the last week. In the
future, it will probably suffice to have the iLab available during the
last week only. During the first 4 weeks, the students concentrated on
getting the simulations working in Matlab.
2. A BIG thank you to Joel for his support. We had continuous
usage during the last week and Joel was online most of the time to help
the students and to restart the iLab server when required. We also had
problems beyond our control (the xPC server crashed from time to time and
needed reboots).
3. The physical hardware held up against several thousands of run. This
is particularly pleasing - the whole class this year ran on only one
inverted pendulum, and this pendulum is still in good condition. In
addition we have 5 others as backup. If we set up a bank of these 6 pendulas,
we could probably serve a total of 600 students at any given time (100 per
pendulum). Any more than this, and the batch queues would probably become
too long. (In comparison, last year we wore out the motor gearbox after only
400 runs).
4. A minimum of three people is required to look after such a setup:
A: An iLab/IT support person (read Joel), B: A technical support person to
maintain the physical hardware locally and C: a lecturer or tutor to
answer experiment-specific questions from the students.
best regards,
Geir