Edge Hardware and Deployment
17 articles
Choosing a Camera for Computer Vision Projects
How to choose a camera for computer vision: USB, CSI, IP/RTSP and depth cameras compared, with advice on shutter, lens, lighting and what really matters.
Dockerize a Computer Vision App for Edge and Cloud
Dockerize a computer vision app: Dockerfile patterns for OpenCV and ONNX Runtime, GPU access, smaller images, pinned versions and reproducible builds.
Export YOLO to ONNX and Run It with ONNX Runtime
Export a YOLO model to ONNX with Ultralytics and run it on CPU with ONNX Runtime, including pre-processing, output decoding and NMS differences.
Hardware Video Decoding: Why Decode Is Often the Real Bottleneck
Every stream is decoded at full resolution before the model sees it. Why decoding can cost more than inference, and how to use hardware decoders.
INT8 Quantization for Detectors: Benefits and Risks
INT8 quantization for detectors: post-training vs quantization-aware training, calibration data, what accuracy risks to expect and how to test them.
Measuring FPS and Latency Correctly
Measure FPS and latency correctly for vision models: warm-up, model time vs end-to-end time, batch effects, GPU synchronisation and a simple benchmark.
Memory Budgets on Small Boards: What Actually Fits
On a Pi or Jetson, memory often runs out before compute. Where it goes, from frame queues to runtimes and the model, and how to measure and cut it.
Microchip Completes Its Hailo Acquisition: What It Means for Edge Vision Projects
Microchip closed its acquisition of edge AI chipmaker Hailo in September 2026. What is known, what is not, and how to manage accelerator vendor risk.
Running Object Detection in the Browser with ONNX Runtime Web
Run object detection in the browser with ONNX Runtime Web: how WASM and WebGPU execution works, model size limits, pre-processing and privacy benefits.
Power, Cooling and Storage for a Reliable Edge AI Box
How to build a reliable edge AI box: power supply quality, thermal throttling and cooling, SD card wear versus SSD storage, and uptime practices.
Raspberry Pi AI Kit and Hailo Accelerators: Is It Worth It?
Is the Raspberry Pi AI Kit with a Hailo accelerator worth it? What it changes for detection projects, the trade-offs, and when a plain CPU is enough.
Raspberry Pi vs Jetson Orin Nano for Object Detection
Raspberry Pi vs Jetson Orin Nano for object detection: how GPU, software stack, power, price and ecosystem should decide which board fits your project.
TensorRT for YOLO: When and How to Use It
TensorRT for YOLO explained: what it optimises, how to build an engine, FP16 and INT8 modes, portability limits and a safe workflow for NVIDIA GPUs.
Thermal Throttling on Edge Devices: Detecting It and Planning for It
Edge boards slow down when they get hot. How to see throttling on a Raspberry Pi or Jetson, how to test for it properly and how to leave enough headroom.
Updating Models on Edge Devices Without Bricking Them
Pushing a new model to boxes you cannot reach is risky. Atomic switches, self-tests, staged rollouts and automatic rollback make updates safe and boring.
Watchdogs and Self-Recovery for Unattended Edge Boxes
An edge box nobody can reach must recover on its own. Hardware watchdogs, systemd restarts, application health checks and remote alerts, layered.
Running YOLO on a Raspberry Pi 5: What to Expect
Running YOLO on a Raspberry Pi 5: setup, model choices, export formats, thread tuning and realistic expectations, with a way to measure speed yourself.