What it does
Our AI-powered mobile robot detects, collects, and transports poultry waste in free-range farms. It solves labour-intensive, unhygienic manual cleaning by automating waste management, improving sanitation and supporting sustainable farming.
Your inspiration
We chose to solve this problem after recognising the major challenges faced by small and medium-sized poultry farms in Malaysia, where manual waste management is time-consuming, labour-intensive, and inconsistent. We observed hygiene issues, odour problems, and health risks linked to poor waste handling. Most existing automated solutions are designed for large, caged farms and are too costly or impractical for smaller operations. This gap inspired us to develop a low-cost, AI-powered mobile robot tailored for free-range environments, making automation accessible, improving hygiene, and reducing labour reliance.
How it works
Our robot is a smart, mobile machine designed to clean poultry farms automatically. It uses a small camera to scan the ground and identify poultry waste with the help of artificial intelligence . Once waste is spotted, the robot moves toward it using motor-powered wheels. A shovel at the front, controlled by a servo motor, then scoops up the waste. The waste is placed onto a conveyor belt, which carries it into a storage bin at the back of the robot. The robot is powered by a rechargeable battery and is controlled by a small computer (Raspberry Pi) that handles all decision-making. The system includes sensors that help the robot avoid obstacles and ensure it does not collide with chickens. All the components are mounted on a lightweight aluminum frame. By combining camera vision, waste detection, smooth motion and mechanical scooping, the robot can clean the farm with minimal human help — reducing labour, improving hygiene and supporting better farm management.
Design process
The design process began by addressing a key issue in poultry farms: manual waste cleaning is slow, unhygienic, and labour-intensive. Our initial concept — a tapping knife mechanism — proved ineffective on uneven terrain during testing. We then redesigned the system using a servo-powered shovel for better scooping and upgraded the conveyor to a dual-chain drive with guiding paddles to prevent slippage. Early 3D-printed PLA parts allowed rapid prototyping but lacked strength, so we switched to a durable ABS plastic shovel. To automate the process, we trained an AI model to detect poultry waste via a mounted camera, with a Raspberry Pi controlling movement and actuation. Field tests on sandy terrain helped us refine the design, improving scooping, stability, and power efficiency — leading to a functional and autonomous prototype.
How it is different
Our design is unique because it is specifically built for free-range poultry farms, unlike most waste-cleaning robots designed for flat, caged environments. It combines AI-powered waste detection, autonomous movement, and a dual-action collection system — a servo-controlled shovel and dual-chain conveyor — into a single compact robot. It can navigate uneven terrain, detect waste using a camera and trained AI model, and scoop and transport it without human input. The system uses modular, low-cost components like a Raspberry Pi and off-the-shelf materials, making it more affordable and practical for small and medium-sized farms.
Future plans
The next step is to improve the robot’s autonomy by integrating onboard AI hardware like Jetson Nano, removing the need for a connected laptop. We aim to expand the training dataset using real farm images to improve waste detection accuracy. Mechanically, we plan to enhance the durability of components and add features like automatic bin emptying. We also intend to conduct long-term field testing in actual poultry farms to gather performance data. Ultimately, we hope to develop a market-ready version that is affordable, scalable, and adaptable for various farm sizes — helping more farmers automate waste management sustainably.
Awards
N/A
Share this page on