It looks like we’ve got quite a few teams who are going to be launching the ball this year. Hopefully we will be one of them. Our initial tests are quite promising.
Since launching the ball requires some significant accelleration/velocity/momentum there is a potential safety hazard, and it seems like a good idea to share some ideas on what might be done to make a launch mechanism as safe as possible. Here’s a couple that we’ve come up with, and I’m throwing them out for comments additions, etc.
1: Keep high velocity moving parts within the starting configuration of the robot.
2: Reduce the potential for accidental launches by having a “safety” button on the OI, so that the driver/co-driver has to press and hold the safety button, and then pull the launch trigger in order to make the ball launcher fire.
3: Have sensors (IR object detectors, the Sharp GP series for example, but simple switches would do) detect whether the trackball is in place before allowing a launch. (Prevents “dry fires” plus having the ball in place means that it is likely that no one’s head is in the way of the launcher… or even within 20" of it!)
4: Have the system automatically and gradually power down when the control signal is lost (end of match, OI unplugged, etc.). For pneumatics this can be done simply by choosing the correct valve setup. For spring loaded systems this may be more difficult to do automatically, but at the very least a safe and simple unloading process should be designed in to the system.
5: Warning lights on the robot to indicate when the system is “armed”.
6: Your ideas…
And, the usual caveat… even if a team or tech inspector posts something here and you implement it on your robot, the definition of “safe” is still the prudent judgement of the tech inspection team at your regional event. CD is not the rule book! These are just some ideas that might help… because it’s much easier to build safety in to a system than to add it on a Thursday at a regional.
All systems that store energy should release that energy slowly when the robot is powered down, but before anyone has a chance to get onto the field and near the robot. By this I mean dissapate the energy over 2-5 seconds.
This means:
1> Pneumatic systems should dump pressure to atmosphere when powered down.
2> Spring loaded and latched systems should unlatch / disengage and dissapate the energy slowly (over a couple of seconds).
3> Winches should release and allow unspooling (especially if used to deform objects to gain spring tension) slowly.
3> Pinch points need to be minimized (watch those 4 bar arms and sissor lifts) while lifting devices are powered down.
Teams should be able to demonstrate these safety measures to an inspector while in the pits. So while just unlatching springs may be safe on the playing field, it may not be safe in the pits.
I would like to point out that while this is a very good idea, the safety of the robot relies completely on the software not malfunctioning (When does that ever happen… :rolleyes: )
All of you launching teams may want to try to implement some sort of physical failsafe that prevents the catapult from firing unless a ball is present. Maybe something like in the cheesy drawing I attached: A lever that is easy for the ball to push out of the way, but it is otherwise in the way of the catapult arm from firing. This would prevent the catapult from doing a ‘dry fire’ if the main latch fails for some reason.
Also, I think it goes without saying that any edges should be very rounded and preferably most of the hard parts of the catapult mechanism somehow padded such that in the event of it contacting a team member while in motion it doesn’t cause too much damage.
In addition, efforts should be made that the systems be as harmless as possible when the robot is not energized. Perhaps having your spring pack connected to the rest of the system with a carabiner. Non-spring based systems should by default generally be fairly harmless when not powered. Just remember to watch when evacuating the air system of the direction things could move. I have been hit by moving parts before.
I’d suggest putting a couple signs on the robot. Something along the lines of “DANGER, SPRING LOADED MECHANISM. DISCHARGE BEFORE INSPECTION OR SERVICING. KEEP CLEAR AT ALL TIMES.”
Safeties are great, but you should also warn people that the thing they are looking at is, in fact, a launcher and not some other mechanism.
We’re following the Battlebots rules for our launcher. That means hardware stops in position, removed when the match is about to start. All pinch joints will be brightly colored, and we’re having a new position in the pits: Fire Commander, whose job it is to make sure the firing path is clear during testing.
For those teams that may be using pneumatics, please be careful in your design. Far too often, a team will design the control system such that at “power on” some part of the robot will extend as the air pressure builds up. This is usually a big surprise to an inspector or other visitor, who may be checking for something inside the robot. Robots parts should not move until told to do so.
Al, could you please help us figure out how to do that? Would it be sensible to use the magnetic sensors for the cylinders to “tell” the RC what position they are in, and switch the solenoid valve to that position at power on?
For those teams that are launching from the floor with those nice big arching trajectories, The amount of potential energy you are storing is not safe. The whole mechanism Is un-safe. Nothing you can do will make it “safe”. All you can do is implement strategies that minimize the risk of injury if the mechanism fails. The question is can you lower the risk of injury to the point that you and the robot inspectors have a subjective opinion that the risk is acceptable.
Jim,
It is usually a simple matter of knowing how the solenoid reacts at power on and making that default position be the “parked” position for actuators. Many times, teams do not take the time to distiguish that. If the action does not cause a robot part to extend out of the sizing box no change is made in the design process. Parts should not move until enabled.
That will work if we make sure to always “park” the mechanism before powering on the robot.
One concern we have with this concept is that this would cause the mechanism to park itself at power off, which could cause unexpected actions for those not familiar with the robot. It would be nice if the robot would just sit there when we kill the power.
Our team plans on launching the ball. To help with safety in the pits and on the field we plan to make a safety latch that goes over our mechanism to prevent injury. We also plan release the energy in the system after the plate has been on without dry firing it.
The main thing that will help stop accidents in the pits and on the field is the fact that everyone working on the robot or in close proxcimity to it should know that it could potentially go off and take every action to ensure that this does not happen. This involves knowing where your team members are and making sure that innocent bystanders are told they need to get away from your launching device so they do not get hurt. You can tell them to admire at a distance until you are sure the system is safe.
We are taking a different approach. As Gdeaver says, a mechanism that can store energy and quickly release is a dangerous thing, and we felt that relying on interlocks and safety procedures is not adequate. So, we designed our mechanism to behave like any other pneumatic actuated device when the ball is not loaded into it (ie. it moves slowly), and to act like a ball launcher only when the ball is in place.
I think it would be difficult to do this with any mechanism that uses springs.
And any mechanism that has a latch (or two or three or…) that suddenly releases a spring loaded bar is intrinsically dangerous.
I’d like to add that any team that is considering a launching mechanism should make sure that they have proper practice as to working the launcher. Like it was said above, the last thing that you want is for your launcher to go off when an inspector, team member, or bystander is too close.
That being said, I would like to remind everyone to try your best to control not only when you launch the ball, but where. Obviously, no one should (and I hope no one would) ever try to aim at another robot. First off it is a penalty, and second it just unsportsmanlike. While accidents happen, please try your best to minimize them by launching so that another robot will not be damaged.
IMHO safety features on the robot during competition only slow us down…or do they? personally, and not to sound arrogant or “holier-than-thou-art”, but in my experience stuff like needing to press 2 buttons to fire, power-down procedures, etc. are over redundant and not needed. when we demonstrate the robot anywhere, yes, we do have as many safety systems as possible, but not usually during the competiton…just my 2 cents…
PS
i guess i’m refering more to having a killswitch and stuff more like that, that is, software safety features. Hardware safety features are a must though
To add to this: Please only use only single solenoid values on your firing mechanism, which default to “not fire”. Double solenoid values stay in their last position, which may not be visually apparent when there is no pressure.
Also, use a momentary push button to launch instead of a several state switch. To reiterate what Jason said some sort of ball sensor should be part of the firing sequence.
Those are my suggestions, here are my concerns:
Please design your robot so that its firing mechanism is not loaded in the Starting Configuration, since the robot must be sized in this configuration without additional restraints. Sliding the robot into the sizing box and using the magic sizing wand on a loaded robot could set it off :ahh: and that is not safe. There will be plenty of time between match start and first hurdle to load you mechanism (the whole 15 sec hybrid mode for instance).
Finally, this is a high speed collision game, so consider 20fps collisions as “normal operating conditions”
Ah, all well and good. But here’s what really concerns me (and if there is already an active thread on this please point me to it)…
This whole IR board thing and what we are allowed to do with it. I personally will be very afraid in the pits if anyone near me is using the same IR channel and accidentally sets off something on my robot or a robot near me.
The IR board is outside the limits of the tether and in my mind that poses a huge safety issue in the pits.
One of my main concerns with launchers is the possiblilty of launching a ball onto another bot. My teams bot should be robust enough to withstand an impact but some bots won’t.
I too am concerned about this. There are some ways to avoid it, including:
Robot OFF when in the pits, unless you are testing.
Robot DISABLED when on, unless you are actually on the field or doing something with it.
Make something that can cover your IR board and shield it while in the pits, unless you need to test it. FLL teams have been encouraged to do this for a while now; a common trick is to line a box with aluminum foil.
*]Warn surrounding teams that you are testing. This gives them time to avoid issues or warn you that there may be an issue.
