agribot uses humidity, dew duration and actual rainfall to flag foliar disease pressure before lesions appear and attach official prevention guidance. [1]
Verification: README documented; not independently verified. These source claims are documented, not independently tested results.
The agricultural problem
Foliar disease pressure can develop before lesions appear; agribot's index uses leaf-wetness and rain-splash pathways to flag that risk. [1]
What the AI agent does
Google Gemini interprets computed disease-pressure numbers and grounds its advice in an official agricultural knowledge base rather than inventing the numbers. [3]
Expected output
The Telegram bot can send a proactive hourly alert for elevated disease risk. [4]
Inputs and data
- The documented data feeds include AgriWeather IoT soil and air sensor readings via its official API and CWA live station observations, including rainfall. [2]
Workflow
- Converse with the Gemini interpreter about deterministically computed disease-pressure numbers and advice grounded in the official agricultural knowledge base. [3]
- Illustrative example: A grower checks whether humidity and dew duration flag foliar disease risk, then reads the official prevention guidance attached to the flag. [1]
- Inspect the Telegram bot's proactive hourly alert for elevated disease risk. [4]
Tools and setup
- The agent runs as a Docker container on a Synology NAS and is delivered through Telegram. [5]
Costs
Unknown — the reviewed source does not verify a cost for this workflow.
Limitations
- Treat thresholds and crop parameters as heuristics calibrated to one specific garden and microclimate, not as settings for a general-purpose product. [6]
Source evidence
Documentation-backed source excerpts, not independent tests of the product. Expand a source to inspect its exact recorded excerpt.
Source [1]DocumentedAgronomic reasoning (deterministic core) - GDD engine — multi-crop growing-degree-day accumulation with per-crop base/upper…
Agronomic reasoning (deterministic core) • GDD engine — multi-crop growing-degree-day accumulation with per-crop base/upper temperatures. • Water demand — ET₀ via FAO-56 Penman–Monteith, converted to crop water demand (ETc) through a growth-stage crop coefficient (Kc). • Disease early-warning — a two-pathway "disease pressure" index (leaf-wetness from humidity/dew duration; rain-splash from actual rainfall) that flags foliar disease risk before lesions appear, then deterministically attaches official prevention guidance from the knowledge base.
Open original documentation ↗Source [2]DocumentedSensing & data - Reads soil temperature / moisture / EC and air temperature / humidity from 阿龜微氣候 (AgriWeather) IoT sensors…
Sensing & data • Reads soil temperature / moisture / EC and air temperature / humidity from 阿龜微氣候 (AgriWeather) IoT sensors via the official API, plus the platform's native irrigation/fertilization advice via an isolated Playwright path. • Pulls 7-day forecasts and live station observations (including rainfall) from Taiwan's Central Weather Administration (CWA).
Open original documentation ↗Source [3]Documentedagribot pairs a deterministic science layer with an LLM interpreter: physical quantities — growing degree days, reference…
agribot pairs a deterministic science layer with an LLM interpreter: physical quantities — growing degree days, reference evapotranspiration, disease pressure — are computed by pure, auditable functions, and Google Gemini is used only to interpret those numbers, converse, and ground its advice in an official agricultural knowledge base — never to invent the numbers itself.
Open original documentation ↗Source [4]DocumentedDelivery - A Telegram bot with scheduled daily briefings, an hourly safety sentinel (proactive alerts for drought /…
Delivery • A Telegram bot with scheduled daily briefings, an hourly safety sentinel (proactive alerts for drought / waterlogging / salinity / elevated disease risk), local commands, and multimodal photo diagnosis. Built on Google Gemini with an agent tool-set and a ReAct loop.
Open original documentation ↗Source [5]DocumentedAn autonomous agricultural monitoring agent for a single vegetable garden, delivered through Telegram. It runs as a Docker…
An autonomous agricultural monitoring agent for a single vegetable garden, delivered through Telegram. It runs as a Docker container on a Synology NAS and watches a leafy-greens plot in the humid subtropical microclimate of Wenshan District, Taipei (near National Chengchi University).
Open original documentation ↗Source [6]DocumentedThis is a personal system built for one specific garden and microclimate, not a general-purpose product. Thresholds and crop…
This is a personal system built for one specific garden and microclimate, not a general-purpose product. Thresholds and crop parameters are calibrated to that context and are honestly flagged as heuristics. It's shared as a reference for anyone building similar deterministic-science-plus-LLM agricultural tooling.
Open original documentation ↗