CropSense
Irrigation Advice From the Soil, in the Farmer's Language
A concept study: we worked up this product, its architecture and its screens as an example of how we would approach it. The screens use fictional demo data and are design concepts, not screenshots of a live product, and this is not a client project.

About the Project
CropSense is a design concept for an irrigation advisory product for growers of vegetables, fruit, sugarcane and cotton. Soil-moisture probes sit at two or three depths in a field and send small readings over a low-power radio network (LoRaWAN) to a gateway on the farm. The gateway passes them to the cloud over a mobile data link.
A water balance model compares those readings with the weather forecast and the growth stage of the crop. Rules set by an agronomist for each crop and soil turn the result into one plain line: irrigate today for a set time, wait, or check the probe. The farmer gets it as an app message, a voice message in their language, a WhatsApp message or an SMS, with the reason in simple words.
A person stays in charge. The agronomist sees every farm on one morning list and can approve, edit or hold advice before it goes out. CropSense is decision support for irrigation timing. It does not replace an agronomist, and it makes no claim about yield or savings.
Why This Matters Now
Market, rule and technology context for this kind of product, checked on 16 November 2026. Each item names its source.
- Swarajya, citing a Moneycontrol report, said India's Digital Agriculture Mission created 11 crore Farmer IDs by late October 2026, about five months before the 31 March 2027 target.Swarajya (news), 24 Oct 2026
- A Press Information Bureau explainer described Kisan Sarathi, a government advisory platform with live expert chat in 13 regional languages, reachable by call centre, WhatsApp, web and app, with IMD and Bhashini integrated.Press Information Bureau (Government of India), 29 Jun 2026
- Rural Voice reported that IMD cut its 2026 monsoon forecast to 90 percent of the long-period average and that the chance of El Nino was expected to be very high through the season.Rural Voice (trade press), 29 May 2026
- The Tribune reported that Agriculture Minister Shivraj Singh Chouhan launched Bharat-VISTAAR, an AI platform for farmer advice built on AgriStack, IMD weather, market prices, pest alerts and ICAR inputs.The Tribune (news), 17 Feb 2026
Requirements & Objectives
- Register a field with its crop, sowing date, soil type and probe positions, by a farmer or a dealer walking the field with a phone.
- Take readings from probes at several depths through a LoRaWAN gateway, and keep working from the gateway's buffer when the mobile link drops.
- Estimate the water available to the crop from probe readings, forecast rain, temperature and crop stage, using rules an agronomist sets for each crop and soil.
- Show each farmer one daily card: irrigate, wait or check, with a duration and the reason in simple words.
- Send advice as an app message, voice playback, a WhatsApp message or an SMS, in the farmer's language, from a fixed template that is checked against the numbers.
- Let an agronomist approve, edit or hold advice from one morning list, and keep every change with a name and time.
- Raise a clear alert to the dealer for a low battery, a stuck reading, a probe out of the soil or a gateway that is offline.
- Cache the last advice on the phone and queue the farmer's Done confirmation until the signal returns.
The Engineered Solution
1Advice built from the field's own soil
Readings at 15, 30 and 45 cm are combined with the forecast and the crop stage, so the advice reflects the root zone of one field and not a regional average.
2The language is part of the product
Each message starts as a fixed set of facts, becomes short sentences with the reason, and is read aloud on request. Native speakers review each language pack before release.
3A person checks the risky advice
The morning list puts advice with low confidence or a flagged probe first. The agronomist can change the duration, hold it or approve it with a note.
4Built for weak signal and low-cost phones
The Android app stays small, keeps the last advice on the phone and queues confirmations. The same advice can go out by SMS or WhatsApp when the app is not in use.
Design Concept Screens
Design concepts with fictional demo data. They show how the product could look and work, not a live release.
01 / 05 Farmer home - read or listen to today's advice in Marathi, see why, and tap Done after irrigating
Engineering Process
- Walk real fields with growers, agronomists and dealers to see how irrigation is decided today and which crops and soils come first.
- Prototype the farmer card, the morning list and the rules editor, and test the wording with growers who prefer to listen.
- Build the probe and gateway pipeline with device identity, a simulator and a database for time-series readings.
- Model the water balance for two crops, write the rules with an agronomist and compare each estimate with how the soil responds after irrigation.
- Create the Android app in two languages with voice playback and offline queuing, then add the SMS and WhatsApp channels.
- Pilot on a small group of farms with agronomist review of all advice, and fix install, battery and signal problems found in the field.
- Harden privacy flows, monitoring and runbooks before widening the pilot to more farms and crops.
Production Tech Stack
| Layer / Component | Production Technologies |
|---|---|
| Farmer app | Kotlin and Jetpack ComposeRoomWorkManagerAndroid text-to-speechFirebase Cloud Messaging |
| Agronomist dashboard | Next.jsTypeScriptTailwind CSSshadcn/uiApache EChartsTanStack Tablenext-intl |
| Field network | Capacitive soil probessolar nodesLoRaWAN gateways with cellular backhaulChirpStackMQTT |
| Backend | Python FastAPI for readings and water balanceNestJS APITemporal for the morning advice run |
| Data | PostgreSQL with TimescaleDB and PostGISRedisobject storage for install photos |
| Languages and weather | Bhashini for translation and speechOpen-Meteo forecastsWhatsApp Business PlatformSMS gatewayAWS Mumbai region |
What the design aims for
CropSense aims to time each irrigation from what a field's own soil is doing, and to explain it in the grower's language, with an agronomist able to review it first.







