I would personally like to see a little independent, remote unit that attempts to regulate something (anything) and reports back as to its success. This would consist of a battery and solar panel for power, RF link, a solid state relay, and some sort of interchangeable sensor. Sort of like a thermostat, but for remote applications in the back yard or warehouse. For example, if I hook it up to a water sensor, and water valve, and give it a source of water to control, like a drip irrigation system, then it would water my plants. The set point would be programmed before deployment or on the front of the unit via a little display and up/down buttons.
But more than that, it would report the status of the battery, sensor readings and relay status to a receiver hooked to my PC. If the drip system gets clogged or the valve gets stuck open, or the battery dies the PC can still see and respond to the water sensor, by alerting someone to the malfunction. Another unit with exactly the same hardware could be connected to a vent actuator or fan on a greenhouse, and use a temperature sensor to keep the growing conditions right and let me know if the vent is not sufficient on hot summer days. Another could run a heater for cold nights. Another could close up the chicken coup when it gets dark at night and open it when the sun comes out in the AM. Dispense food when the feed bowl is empty.
They could be used inside to provide thermostatic control for each room in the house: The unit in each room has a motor to run the vent open or closed. The PC controls the main thermostat. If one room needs heat and the others don't, the PC turns on the heat, and that room opens the vent, while the others close theirs. Installing that system would normally require a wire from a power source to a vent actuator AND a wire from a temp sensor in each room to the central PC. With solar / battery / wireless, you just mount the unit on the vent right next to the vent actuator motor. In most cases, that actuator could easily be a pager motor with a fine thread screw and a rider nut attached to the handle of the vent.
Of course, the solar panel and battery introduce an opportunity for failure: The load you are controlling may drain the battery. Part of the smarts built into the thing has to be changing the hysteresis of the loop to accommodate that issue. E.g. if the water valve really sucks down the battery, the unit will not turn it on until the plants are really getting dry, then will leave it on until they are pretty darn wet. Still better than nothing, and the variation is being reported to the PC so the operator can see that perhaps a larger battery / panel or some other source of power would be best.
The units could, of course, be made to work with a constant power source from the AC or other supply, but the logic for battery power management should be built into the system.