The Challenge
Omlet, a leading pet products company, wanted to offer customers a smart monitoring system for chicken coops—allowing owners to verify their flock's safety remotely and receive alerts when chickens hadn't returned to the coop at night. The challenge was fundamental: how do you run sophisticated AI object detection on a battery-powered camera in a garden shed with no mains power and limited connectivity?
Traditional computer vision approaches require significant computational resources, typically cloud processing or powerful local hardware. Neither option works for remote monitoring—cloud processing demands constant connectivity and introduces latency, while powerful hardware drains batteries in hours. Omlet needed an AI solution that could run continuously on minimal power while maintaining accurate detection in challenging outdoor conditions: variable lighting, weather exposure, and subjects that don't stay still.
Our Approach
New Gradient developed a purpose-built lightweight neural network architecture optimised specifically for edge deployment. Rather than shrinking a standard detection model, we designed our network from the ground up with power efficiency as a primary constraint, selecting operations and layer structures that maximise accuracy per watt consumed.
The training process incorporated the specific challenges of the deployment environment. We collected and annotated thousands of images representing the full range of conditions the system would encounter: dawn and dusk lighting, varying weather, partial occlusions, and multiple chickens in frame simultaneously. The resulting model achieves robust detection across these scenarios while fitting within the tight computational budget of battery-powered edge hardware.
We implemented aggressive optimisation techniques including quantisation, pruning, and efficient inference scheduling. The system intelligently manages when to run inference, balancing responsiveness with battery life. When chickens are detected in the run area at dusk, the system increases monitoring frequency; during quiet periods, it scales back to conserve power.
The Outcome
The deployed system delivers real-time chicken detection and counting on battery-powered IP cameras, enabling Omlet to offer a genuinely useful smart coop product. Users receive instant mobile notifications when their chickens haven't returned to the coop, along with accurate counts of birds in different areas. The system operates for extended periods on battery power alone, making it practical for typical garden installations without mains electricity.
The solution handles challenging detection scenarios reliably—chickens are notoriously variable in appearance and behaviour, and outdoor conditions change constantly. Our optimised model maintains accuracy across lighting conditions from bright daylight to the low-light periods when monitoring matters most, as chickens return to roost at dusk.
Beyond this specific application, the project demonstrates New Gradient's capability in edge AI deployment. The same principles—purpose-built architectures, aggressive optimisation, and intelligent inference scheduling—apply across IoT and smart device applications where power, connectivity, or latency constraints preclude traditional cloud-based approaches. Wildlife monitoring, agricultural sensing, and remote infrastructure inspection all present similar challenges where edge AI unlocks previously impractical solutions.