Yes, you just need a suitable solenoid valve controlled by a timer. You can build a timer, but I would just buy a programmable timer relay.

 

Search "electric air valve". Small ones are $20.

ak

 

Kind of late for this thread... but my opinion is that, although a mechanical solution is feasible, an electronic one would be nowadays cheaper... just make sure to include hysteresis into the control algorithm, that's the key to the project.
Weight the crickets, and average them for the duration of a few days, then let go of the trap.

 

If you could get them to gather in a tube, e.g. 100 to 150mm plastic drainpipe, you could fire them them out the end, and through a one-way flap valve, into the collection container with a blast of compressed air. That way, they would have no choice in the matter. This method also has the advantages of very few moving parts and the absence of any powered door mechanism means it is unlikely to trap or injure the crickets

Another option might be to use vacuum to suck them into a jar connect the inlet hose of a " shopvac " to a hole in the lid of a fairly large jar ( a gallon or so ) . connect another hose to another hole in the jar's lid that goes to the cricket collection point . When your sensors ( weight , light , capacitive prox , or whatever works ) detects the crickets , a relay would turn on the shop vac for a predetermined time and suck them into the jar . Too much vacuum airflow could suck them through the jar and into the shop-vac . But , maybe this would not be a problem anyway , just suck them into the shop vac and not use the jar at all .

 

<snip> Too much vacuum airflow could suck them through the jar and into the shop-vac . But , maybe this would not be a problem anyway , just suck them into the shop vac and not use the jar at all .

Or use a filter over the shop vac hose (woman's knee high stocking) to prevent anything being sucked into the vac.

I personally like the compressed air option. Then you could mechanically move in separate containers.