this is for holding javascript data
Rein D. Otsason renamed EM Improve - SAP.tex to EM Improve SAP.tex
about 10 years ago
Commit id: c963b7024886090fe2e1d24dfa7cc20eb47c1e4e
deletions | additions
diff --git a/EM Improve SAP.tex b/EM Improve SAP.tex
new file mode 100644
index 0000000..f14af20
--- /dev/null
+++ b/EM Improve SAP.tex
...
\subsubsection{Improving reliability of the switch activation arrangement}
A persistent problem during the final stages of the project was achieving full functionality of the switch activation arrangement.
The causes of issues were:
\begin{enumerate}
\item Poor alignment
\item Variability of the test units
\item Placement of the test units
\end{enumerate}
Problem (1) was the first to be encountered, and also the first to be fixed.
Poor alignment of the test units in the tray with the switch activation plate lead to cases where,
for example, given a motion of the switch activation plate sufficient to move the switch on a test unit into the on position,
only one light would go on, or only one row.
These problems were identified early, and mostly resolved by the end of the project,
though the process would have been greatly simplified through more precise manufacturing techniques.
Problem (2) was more difficult to resolve.
The initial position of the switches on the lights,
the required distance to activate, and the required force
all varied between units to a significant extent.
The selected design relied on precise manufacture of grooves to interface with the switches,
and when a similar level of precision was lacking in the test units, the mechanism's performance was sub-optimal.
This problem could be corrected by cutting more generous grooves into the switch activation plate,
and filling the grooves with some sort of pliable material to allow a more flexible activation of the lights.
Problem (3) arose late in the project, and would have been easier to solve if only more time had been available.
Interaction with the client clarified the meaning of an ambiguous term in the RFP.
This necessitated the removal of padding which prevented the rotation and vertical movement of the test units via friction.
The restriction of vertical motion was most critical, since without some sort of downwards force
the test units would move upwards without being activated during operation of the switch activation arrangement.
Possible solutions would be some attachment to the sensor array to apply downwards pressure,
or some manner of spring on the tray to hold the lights down after being placed in the tray.
Similar behaviour could be accomplished with a lid on the tray.
