A hydroponic system or a classroom garden can survive a school holiday, but only if you prepare it before you lock the door. The ways it dies are boring and predictable: an empty reservoir, a pump left on a dead timer, a nutrient mix that drifted too strong while nobody was watching. Work through the checklist below in the last lesson before the break and most of those failures disappear.

Top up the reservoir and leave EC headroom

Fill the reservoir to full. Over a two-week break the plants drink and the surface evaporates, so a tank that is only half full at the start can run dry before anyone returns. A bigger volume of water also changes temperature more slowly, which is kinder to roots.

Here is the part people miss: as water leaves, the nutrients left behind get more concentrated, so the EC creeps up. Mix your solution slightly weaker than usual before a long absence. You want the mix to still be in a safe range at the end of the break, not just the start. For soil beds, water deeply once and mulch the surface to slow evaporation. If you run a hydroponic setup, this single step of topping up and dialling the strength down is the most common save.

Test every timer and pump before you trust them

Do not trust a timer you have not watched run. Trigger each cycle by hand, watch the pump kick on, and watch it switch off again. A timer that looks set but never fires will quietly starve the whole system.

In South Africa, load shedding is the extra variable. A power cut resets many cheap mechanical timers and can leave a pump stuck off, or worse, stuck on. Check the load-shedding schedule for the break dates in your area, and know what your equipment does when power comes back. If a pump does not resume on its own, that is a problem you want to design out before you leave, not discover afterwards.

Set up remote alerts you will actually see

The point of automation over a holiday is not to run everything unattended. It is to tell you the moment something breaks. A water-level sensor, a temperature reading, and a simple "pump did not run today" message cover most of what goes wrong.

Automate sensing and alerting first, before you automate anything clever like auto-dosing. Knowing early that a tank is low beats a complicated dosing rig that fails silently. Route the alerts to a phone you actually check, not an inbox you ignore over the holidays. A monitoring platform such as sheenIoT can push a notification when a reading crosses a line, and the same approach works whether the plants are at school or in a home setup. Before you leave, trip one alert on purpose so you know the message really lands.

Name a human fallback

Automation fails. Batteries die, WiFi drops, a seal leaks. So name one person who can physically get to the plants during the break. Hand them a key and a one-line instruction: top the tank up to full and send me a photo. That named human is the single most reliable line in the whole plan, and it costs nothing to arrange.

The pre-break checklist

• Reservoir topped up to full
• Nutrient mix left slightly weak so EC has headroom to climb
• Every timer triggered by hand and watched through a full on-off cycle
• Each pump confirmed running and stopping
• At least one remote alert tested by tripping it on purpose
• Load-shedding schedule checked against the break dates
• One named person with access and a one-line instruction
• A photo of the healthy baseline so you can compare on the day you return

None of this is complicated, and it does not need to be automated to work. Start with the reservoir and the named human, add a sensor and an alert next, and leave for the break knowing the plants will still be there when the term starts again.