AGRINOVA — PURPLE UMBRELLA FESTIVAL 2026 Transcript of the stage pitch — AgriNova Team. Length ~5:50. ================ PART 1 — THE OPENING (clip 2, 2:25) ================ Seventy percent. Seventy percent of the world's fresh water is used for agriculture. Yet small farms still irrigate on guesswork — watering by habit, not by need. Most farmers can't watch their crops overnight. Animals walk in, crops get destroyed, and the farmer only finds out the next day when they come back to check the state of the crops. By the end of the month, that's a lot of wasted water — and lost crops. [next slide] So — what if there was a system that could guard a farm, and water it automatically, for less than 2,500 rupees? [next slide] It's real. And we built it. [next slide] AgriNova is our answer. A farm that texts you back. AgriNova is an open-source project dedicated to helping farmers monitor their farm remotely — wherever they are. We don't use a proprietary app that works only with our system and needs a learning curve. We use Telegram. Telegram is as simple as WhatsApp — simple chatting. The farmer already knows how to use it. ================ PART 2 — HOW IT WORKS TO CLOSE (clip 1, 3:26) ================ Now let me show you how it works. It's built in three parts: the reflex, the hands, and the brain. THE REFLEX — the reflex is pure hardware. It watches the IR sensor and it cannot crash, it cannot hang — even if every other computer in the system fails, you still get motion detection from the IR sensor, because the logic is burned into the chip. THE HANDS — the hands are an Arduino microcontroller. It reads all the sensor data every second and packages it into one format for the brain. THE BRAIN — the brain is a Python program. It takes all of this data, processes it, and runs the Telegram bot, the camera, and the graphs. [next slide] Now let's look at the numbers. It's built for the smallest farmers: about 2,500 rupees — a fraction of the cost of commercial systems. Under 10 seconds from something entering your farm to the alert, photo and video being sent to your phone. And around 67 percent less water — that number is PREDICTED; the field trial is our next step, and after we've performed the field trial we'll have the complete number of how much water is saved. There's no app, there's no subscription, and it talks to you in English, Hindi and Telugu. [next slide] You might ask — where is the AI? This year's theme is AI and Climate. We chose not to use it. The responsible use of AI is knowing when not to use it. AI data centres are not eco-friendly, and farms don't have a stable internet connection to keep a connection to a server. We could use on-device AI, but it doesn't improve things significantly here. Farm decisions are time-critical — we can't wait for a model to answer and then check it for hallucinations. That's why there's no AI in the system. [next slide] Now let's take a look at the physical prototype. This is the front view of our model. You can see the camera on the pole, so the farmer can get a photo or video of the farm. We've also included a solar panel — if wall power isn't available, it can switch to solar electricity. And our sensor cluster, which includes all the sensors needed to send the data back: the IR sensor, the temperature-and-humidity sensor, and the soil-moisture sensor. In the side view — the blue board is the Arduino UNO microcontroller, the bigger board is the FPGA controller, and the thing that looks a bit like a standard switch is our mist pump with its relay. And in the top view you can see the sensors and the wiring over the two farm beds. [next slide] That's AgriNova. Do you have any questions I can answer?