pre-charged Pneumatics

[Bill_B]

Aircraft usually have a collection of locking pins with attached flags saying "remove before flight". They extend into visibility so that the ground crew can prepare the plane for use.

[Tristan Lall]

Originally Posted by Al Skierkiewicz

Peter and Howie,
This is the rule that applies...

<R90> The ROBOT will be inspected for compliance with the dimension constraints specified in Rule <R10> while in its NORMAL CONFIGURATION, by being placed within a FIRST Sizing Device that has inside surface dimensions consistent with the rule. Other than resting on the floor of the Sizing Device, no part of the ROBOT can break the plane of the sides or top of the Sizing Device during size inspection. The ROBOT must be self-supporting while in the Sizing Device.

Inspections are carried out with all systems unpressurized and unpowered for obvious safety reasons. Robot size and weight are tested with the bumpers off and the battery out. Please refer to Rule <10> and <11> as both are referenced in robot size.

The rules for stored energy govern the condition that a robot can be in at the start of a match including air stored in the Clippards. This rule allows teams to pressurize their robot with an off robot compressor but speaks to the start of a match, not to inspection. Please ask the question as we have no guidance to the contrary from the GDC. I expect the pressure vent valve to be open until the "power on" portion of the inspection takes place.

Al, I think I also need to disagree a little here. As the rules currently stand, normal configuration definitely implies a powered-on, disabled state ("the state of the ROBOT immediately before being enabled by the Field Management System, before the ROBOT takes any actions, deploys any mechanisms, or moves away from the starting location"). Although for reasons of practicality and safety it would seem valid to inspect robots in their least-energy state, if doing so would inaccurately simulate the normal configuration during sizing, the rules would seem to demand that it be tested in the normal configuration (per <R88> and <R90>).

Of course, there are not many mechanisms for which this is a valid concern, because with control system outputs disabled, control of most actuators and other devices is not practical. One possibility that comes to mind is an electromagnet that receives power directly from the PDB, through a mechanical switch, and which holds down some component until that switch is triggered, cutting power. (This is a custom circuit, and by using a non-relay switch, avoids being controlled by the cRIO.)

Now, it is certainly within an inspector's authority—and indeed, responsibility—to take steps to minimize the risk to themself and others during inspection. In the above case, if the team were to argue that they needed that component to be held back, but the inspector was uncomfortable with the presence of an armed mechanism in the sizing box, perhaps a compromise could be reached where the mechanism is fixed in the appropriate position, but in a locked-out state.

But in other cases, if the inspector considers the hazard relatively minor, I don't think that the rules would prevent a team from asking to be sized in a powered-on, disabled state.

[Al Skierkiewicz]

Tristan,
As I have pointed out in a different thread, we have received no direction from the GDC on any changes to the method of inspection. I cannot agree, based on past safety discussions and procedures, that transporting a robot with a charged mechanism, pneumatic or mechanical, is appropriate. If that is the case, then inspecting under different conditions is also inappropriate. Teams for many years have figured out ways to move robot objects into playing position that did not require charged systems to do so.
I think that the majority of this discussion likely is directed at kicking mechanisms. If the kicking device would fall outside of the robot frame perimeter without being in the charged state, it seems to me that the team needs to design their mechanism to keep the kicker inside the frame perimeter when at rest.

[Tristan Lall]

Originally Posted by Al Skierkiewicz

Tristan,
As I have pointed out in a different thread, we have received no direction from the GDC on any changes to the method of inspection. I cannot agree, based on past safety discussions and procedures, that transporting a robot with a charged mechanism, pneumatic or mechanical, is appropriate. If that is the case, then inspecting under different conditions is also inappropriate. Teams for many years have figured out ways to move robot objects into playing position that did not require charged systems to do so.

That's not exactly the point I was making; basically, there's a balance to be struck between a team's possible expectation (following directly from the rulebook) that the robot will be inspected in the normal configuration (powered, disabled), and the inspectors' need to perform a safe inspection.

If some aspect of a robot is designed to comply with the size limits only when powered (i.e. in normal configuration), and it can be inspected safely while in this state, then there's no need to categorically insist that this particular robot be unpowered during inspection. But by the same token, if the inspector cannot safely perform an inspection while this particular robot is powered, then there's no question that he has the right to instruct the team to make their robot safe while still accurately simulating the normal configuration. (For many robots, an accurate simulation of normal configuration could be achieved with power off, pneumatics discharged, and springs at rest position—they're not the ones I'm concerned about.)

The exact steps necessary to achieve this will vary from robot to robot, but might, for example, include disconnecting the cRIO and all nonessential circuit breakers, while still allowing the custom circuit to remain powered. That's totally reasonable for something like a lifting hook that is retained electromagnetically, but which doesn't pose much of a risk. That's not necessarily sufficient for something like an electromagnetically-retained ball kicker, because of significant amounts of energy stored in the system.

In principle, I think the safety lock system envisioned by eagle33199 would be another reasonable way to address this (obviously provided that the lockout method was adequate). The lockout doesn't have to be elaborate, just effective: for low-energy systems, it might be a zip tie in place of an electromagnet, for higher-energy systems, it might be a well-placed two-by-four physically stopping the kicker from reaching its outermost limit.

[Al Skierkiewicz]

Tristan,
Weight and size is performed with the bumpers off and the battery out. Inspection continues with no power until such time as the inspector reaches the "power on tests" at the end of the inspection.
As Mike has pointed out in example above, there will be mechanisms that require locking pins to transport the robot safely onto the field. I agree with his assessment. If removal of the lock cannot keep the robot within the sizing box, then it doesn't pass the test of "self-supporting" in <R90>. If the robot is prepared for a match by storing energy in one of the allowed methods in <R01> and the locking device makes handling and transport safe then as Mike has put it, it's use should be mandatory.
Again, until the GDC releases a rule governing this situation or the inspection checklist is released, this is my opinion and no one else. In training inspectors, I cannot advise them to inspect a robot powered, precharged, or predeflected until such time as there is a need to demonstrate the power on tests or the energy and travel of a moving robot mechanism as part of that inspection.

[Mike Betts]

Tristan,

Concerning electromagnets, the GDC has been slow to make a definitive, clear ruling. As Al put it, what follows is my personal viewpoint...

A very popular electric solenoid actuator is a hinged clapper design. This is a ferrous core electromagnet with a spring loaded ferrous plate hinged at one end. Applying power to the electromagnet causes the plate to clap to the magnet and deflects the spring. When the EM is de-energized, the spring moves the clapper away from the EM.

Electric solenoid actuators are not allowed as per <R53>.

Many of the designs I have read about sound like the team has made their own ESA and are calling it a magnetic lock. This would not be allowed under this year's rules.

Concerning powered up robots. Al is 100% correct. The beginning of the inspection process is done with the battery removed...

My last point is safety. The idea that unpowering a robot, intentionally or unintentionally that causes a mechanism to deploy is just not safe.

Bottom line: While EMs are being allowed by the GDC, I can not envision an non-trivial, EM system that could be legally used on this year's robot.

IMHO, the GDC has dug a hole and will have to enlighten us further...

Regards,

Mike

[Tristan Lall]

Originally Posted by Al Skierkiewicz

Tristan,
Weight and size is performed with the bumpers off and the battery out. Inspection continues with no power until such time as the inspector reaches the "power on tests" at the end of the inspection.
As Mike has pointed out in example above, there will be mechanisms that require locking pins to transport the robot safely onto the field. I agree with his assessment. If removal of the lock cannot keep the robot within the sizing box, then it doesn't pass the test of "self-supporting" in <R90>. If the robot is prepared for a match by storing energy in one of the allowed methods in <R01> and the locking device makes handling and transport safe then as Mike has put it, it's use should be mandatory.
Again, until the GDC releases a rule governing this situation or the inspection checklist is released, this is my opinion and no one else. In training inspectors, I cannot advise them to inspect a robot powered, precharged, or predeflected until such time as there is a need to demonstrate the power on tests or the energy and travel of a moving robot mechanism as part of that inspection.

Let me put this another way—the issue is with this year's rules:

Originally Posted by Section 8.2

NORMAL CONFIGURATION – The physical configuration and orientation of the ROBOT when the MATCH is started. This is the state of the ROBOT immediately before being enabled by the Field Management System, before the ROBOT takes any actions, deploys any mechanisms, or moves away from the starting location. This configuration is static, and does not change during a single MATCH (although it may change from MATCH to MATCH).

Originally Posted by <R90>

The ROBOT will be inspected for compliance with the dimension constraints specified in Rule <R10> while in its NORMAL CONFIGURATION, by being placed within a FIRST Sizing Device that has inside surface dimensions consistent with the rule. Other than resting on the floor of the Sizing Device, no part of the ROBOT can break the plane of the sides or top of the Sizing Device during size inspection. The ROBOT must be self-supporting while in the Sizing Device.

It is explicitly stated that size inspection takes place while in normal configuration. The most straightforward reading of "immediately before being enabled by the [FMS]" leads me to believe that they are defining normal configuration in terms of the instant before a match begins—the robot is powered, and disabled. Therefore, despite usual practice (of which I am certainly aware), and perhaps our better judgment, the 2010 rules do not specify size inspection with "the battery out".

If a team reads <R90> and constructs a mechanism that is self-supporting in normal configuration, but either non-self-supporting or outside the size limits when unpowered, and then presents it for inspection, the inspectors need to formulate an appropriate way to measure its size, while avoiding the imposition of ad hoc constraints and simultaneously maintaining a safe process.

If we leave that particular robot unpowered and unrestrained, it will fail—but these conditions are inconsistent with normal configuration. This is therefore an unacceptable basis upon which to evaluate rules compliance for that robot. (For other robots, where the unpowered configuration is clearly equivalent to the normal configuration, this is moot.)

On the other hand, if we power it and leave it unrestrained, there may be unnecessary hazards to the inspectors and bystanders. Only if the risk is minimal would this be appropriate.

If the risk of harm is too high in the previous scenario, then we must consider restraining it. If we're willing to interpret the self-supporting requirement in <R90> as a constraint on design (in other words, the robot must be able to be self-supporting in the box), rather than as a constraint on the precise process used by the inspectors, then it would be sufficient for the team to independently demonstrate self-containment, and then put it in the box with a restraint to permit it to fit and to preserve safety. (For example, a robot's kicker is known to be self-contained when in normal configuration, but requires power. However, the robot will not fit with the kicker extended. Partially retract the kicker and restrain it securely, so that it fits in the box. Check for size.)

Alternatively, if the shape of the sizing box permits it, and the robot had already demonstrated the capability to self-constrain when powered, you could simply insert the unpowered, unrestrained robot and ignore the protrusion.


Now, with regard to electromagnets: I presented that as an example of a mechanism that would require power to stay within size limits, but which did not depend on the controls being enabled. If that's indeed ruled illegal, then for the purposes of my example, we can imagine another mechanism that behaves similarly.

As an aside, I'm not convinced that that clapper mechanism is a solenoid actuator—it seems to fail the usual definition of a solenoid (an electromagnetic coil surrounding a cavity, within which there is a uniform magnetic field) or a solenoid actuator (which moves an armature linearly within the cavity). It is definitely an electromagnetic actuator—but if there was no intent to distinguish a solenoid from any other electromagnet, why use the specific term?

Similarly, with the electromagnetic locks, their geometry is totally different from a solenoid—they rely on the field outside of the coil, and do not directly drive an armature.

[Mike Betts]

You know, I think that our first Lead Inspector conference call is going to be very interesting...