Build a Raspberry Pi Beehive Monitor: Load-Cell Hive Scale, Environmental Sensors, and Camera-Based Activity Tracking

A hive scale is one of the most information-dense tools a beekeeper can add to an apiary, because weight is the single best proxy for what's actually happening inside a closed hive without opening it and disturbing the colony. A sudden multi-pound overnight drop can mean a swarm just left. A steady multi-week climb means a strong nectar flow is coming in and it might be time to add a super. A slow winter decline tells you how much of the colony's stores are left and whether emergency feeding is needed before spring. This project builds a Raspberry Pi-based monitoring station that logs hive weight continuously via load cells under the hive stand, pairs it with temperature, humidity, and barometric pressure data, and adds a camera and simple audio monitoring for entrance activity and swarm-sound detection — all without lifting a single frame.

Load cell placement and the four-corner approach

Rather than a single central load cell, mount a bar-type load cell under each corner of the hive stand platform, wired into a single HX711 amplifier (most 4-cell setups use a combinator board, or sum the four analog outputs through a proper load-cell summing junction box before the single HX711 channel) or into four separate HX711 channels if you want per-corner data. Four corners gives you two things a single center-mounted cell can't: total hive weight that's accurate even if the hive isn't perfectly centered on the platform, and a rough bee-distribution signal — healthy colonies in warm weather often shift weight distribution across the hive body in ways a single-point scale would miss entirely, while a sudden one-corner-only weight change can indicate the hive has started to tip or a stand leg has settled into soft ground.

Build the platform itself from marine plywood or pressure-treated lumber sized to the beekeeper's standard hive footprint, with the load cells sandwiched between the platform and a rigid base resting on level ground or concrete blocks. Get this level before the hive goes on — an unlevel scale platform introduces a consistent per-corner bias that's easy to mistake for a real hive-health signal if you don't zero and calibrate the system with the actual empty equipment in place first.

Calibration

Calibrate with known reference weights (a few full 5-gallon water jugs work well, or a properly weighed deep hive body before bees are added) rather than trusting the HX711's default scale factor, and re-zero whenever you add or remove a super, since each box changes the system's tare weight and you want the logged trend to reflect bee and stores weight, not container weight. Log both raw weight and a clearly-annotated tare event in your data so a graph of "hive weight over the season" isn't confused by the day you added a second brood box.

Weight signalLikely causeBeekeeper action Sudden drop of several pounds within hours, especially midday on a warm daySwarm departureCheck for swarm cells on next inspection; consider splitting proactively next season Steady daily gain of 1-5+ lbs over multiple daysStrong nectar flowCheck super capacity, add space if needed Gradual decline through winterNormal stores consumptionCompare rate against historical data; feed if declining faster than prior years Oscillating daily pattern (up at night, down in day)Normal foraging bee mass leaving/returning plus evaporation of nectar moistureExpected — don't over-react to daily noise, watch the multi-day trend

Environmental sensors and placement

Mount the BME280 in a ventilated, shaded enclosure near the hive but not touching it directly, to capture ambient conditions rather than the hive's own internal microclimate (which runs warmer and more humid than ambient, especially near the brood nest) — ambient weather data is what you want for correlating against nectar flow timing and swarm risk days (swarms disproportionately happen on warm, calm days following a stretch of bad weather that kept bees confined). If you want internal hive temperature specifically — a genuinely useful signal for detecting queenlessness, since a hive without a laying queen loses its tightly-regulated brood-nest temperature — that requires a separate waterproof probe sensor fed through a small hole in the hive body, which is a reasonable future upgrade but adds another point where the hive envelope needs sealing against moisture and robber bees.

Camera and sound: entrance activity without opening the hive

Position the Pi Camera Module to frame the hive entrance from a few feet away, and run a simple time-lapse or motion-triggered capture rather than continuous video recording, which saves storage and bandwidth while still giving you a visual record of entrance traffic level, pollen-carrying forager activity, and — critically — the distinctive visual signature of a swarm boiling out of the entrance, which is unmistakable on video even from a layperson's glance. A small electret microphone near (not inside) the hive can pick up the audible frequency shift bees make when queenless or preparing to swarm; this is a harder signal to use reliably than weight or camera data, but even a simple audio level/frequency logging setup gives experienced beekeepers another data point to correlate against the others.

Power and remote apiary considerations

Many apiaries sit well away from mains power, so budget for a small solar panel and battery bank sized to the Pi's modest draw — a Pi Zero 2 W with a camera doing occasional time-lapse captures and periodic sensor reads is a very light load, and a 20W panel with a modest 12V battery through a buck converter will run it through most weather. Use a long-range WiFi bridge or a cellular modem if the apiary is beyond home WiFi range, and log locally to the Pi's SD card (or better, to a small USB drive, since constant writes are harder on SD cards over a multi-season deployment) with periodic sync whenever connectivity is available, so you don't lose data during connectivity gaps.

Dashboard and alerts

Push weight, temperature, and humidity readings to Home Assistant or a simple Grafana/InfluxDB stack (both already covered elsewhere in this library for other sensor projects) and set up an alert for any weight drop exceeding a threshold within a short window — that's the signal most worth an immediate notification, since a same-day swarm catch is realistic but a swarm discovered a week later during a routine inspection is long gone.

Weatherproof every connection point generously; an apiary enclosure lives outdoors in direct weather for an entire season without the kind of attention a workshop electronics project gets, and a corroded HX711 connector reading garbage weight data is a far more common failure than anything in the firmware.