[Al Skierkiewicz]

OK,
I can answer some of these concerns. Be aware that not all designs will fit all parts of this discussion and the event LRI still has the authority to rule something "unsafe" when they are examining the individual robot. As always, the Q&A is the final authority and the inspection staff will inspect as directed by the GDC and the official Q&A.
As far as launch mechanisms in general, a safe device will only launch the ball and nothing else, EVER! This is the basis for inspectors to see what happens in a dry fire situation. Often a launch mechanism unloaded acts much differently than when a ball is in place. We don't want and cannot allow, a mechanism that appears as if it can launch pieces of itself into people standing near the field or in the stands. Additionally, we do not want to see launch mechanisms that can trigger accidentally anywhere in the competition venue. That holds true for the queue, near or on the playing field, the pit, or during inspection.
As to general areas that are under scrutiny every year...
1. Surgical tubing needs to be securely fastened to the robot so as to not pull out under tension and cause harm to those close by.
2. Springs are of concern under certain uses where shock is caused to the spring during the firing period. This shock may cause cracking or other metal fatigue. A common method in this case is to string aircraft cable through the spring. Should the spring fail, the broken parts will generally be restrained by the cable. COTS springs that are used without modification (i.e. are not cut, ground or heated, etc. by the team) are generally acceptable. A garage door spring (generally not appropriate), cut to length by the team for example, is suspect to fail.
3. Pneumatics when used for launch mechanisms must conform to all pneumatic rules. That means any stored air pressure must be provided by the one and only one robot compressor. Should a system be capable of causing a rise in working pressure as a piston is moved, then the increase in pressure must escape through the required, pressure (Norgren regulator P/N: R07-100-RNEA recommended) relieving primary regulator. At the end of the match, all pressure on the robot must be relieved by the one and only one pressure vent plug valve.
4. Other parts of the mechanism may be subject to considerable stresses during a firing sequence. As such, parts held together with unsuitable fasteners will be questioned.
5. There has been some recent answers in the Q&A that relate to the above, please review and apply as needed in your design.
I can tell you that inspectors are very excited about seeing some of these mechanisms actually play this year. This kind of game really brings out the creativity in design. We want to see you play but we are serious about safe play and operation. Please remember safety when testing your mechanisms at an event especially in the pit.