Arducam recently launched the All-in-One Raspberry Pi AI Camera Kit with CM5, a Raspberry Pi CM5-based PoE outdoor security camera featuring a 12.3MP Sony IMX500 AI vision sensor housed in an IP66-rated enclosure.
The camera combines pixel-level sensing and AI inference on a single chip, reducing the need for cloud processing while improving efficiency and privacy. The security camera supports operation from -20°C to +75°C, although image stability is only guaranteed in the 0 to 50°C range, potentially making it unsuitable for locations where temperatures drop below freezing. It can be used for perimeter security, construction site safety monitoring, traffic and parking enforcement, and public-space crowd analytics.
All-in-One Raspberry Pi AI Camera Kit specifications
- SoM – Raspberry Pi Compute Module 5 (CM5) by default CM5004016 with 4GB RAM, 16GB eMMC Flash, no Wi-Fi and Bluetooth
- Image Sensor
- 12.3MPÂ Sony IMX500 Intelligent Vision Sensor (as found on the Raspberry Pi AI camera)
- Chroma – Color
- Shutter Type – Rolling Shutter
- Optical Size – 1/2.3″
- Pixel Size – 1.55µm x 1.55µm
- Sensor Active area – 4056(H)×3040(V)
- Output Format – RAW10 (Platform Output: JPEG, YUV, RGB)
- Sensor Image Area – 6.29 mm x 4.71 mm (7.86 mm diagonal)
- Video Modes – 4056×3040 @ 10fps, 2028×1520@ 30fps
- Lens
- Fixed Focus
- Focal Length – 4.7 mm
- F.NO – F1.79
- Field of View(FOV) – 78°(D)×66.3°(H)×52.3°(V)
- Lens Mount – Stock Lens Mount
- IR Sensitivity – Integral 650nm IR Filter, visible light only
- Focus Distance 20cm-INF
- Networking
- Ethernet Port with PoE
- Wi-Fi and Bluetooth connectivity are NOT available
- Power
- IEEE 802.3at PoE/50-57V 600mA
- Consumption 15W MAX
- Dimensions – 153.05 x 82.09 x 79.10 mm
- Temperature Range
- Operating: -20°C to +75°C (but image stability is only guaranteed between 0°C and 50°C)
- Storage: -10°C to +75°C
- Rating – Metal enclosure with IP66 (Waterproof, Dustproof, Fogproof)
- Certifications – CE, RoHS
While writing the specification, I found some inconsistencies in the specifications listed on the product page and the datasheet in three areas. The operating temperature is mentioned as -20°C to +75°C on the product page, but -10°C to +75°C in the datasheet. The optics also differ, with the aperture listed as F1.79 versus F1.9, which could be a typo, unless Arducam changed the lens glass for the IP66 housing. Lastly, the products page mentions that the device supports JPEG/YUV/RGB/RAW10 outputs, whereas the datasheet mentions that only the RAW10 output is supported.
The camera runs Raspberry Pi OS and supports the IMX500 AI Model Zoo, which provides more options than the standard Sony AITRIOS templates. You can run pre-trained models for tasks like object detection, pose estimation, and segmentation. Since inference is handled directly on the IMX500 sensor, the CM5 can focus on application logic, networking, and data storage instead of AI processing.
The company has also released a compact version of the camera (B0581), with the same Sony IMX500 vision sensor, a small enclosure, and USB-C or PoE power options. In comparison, the Outdoor Security AI Camera Kit (B0572) is designed for outdoor deployments, with a significantly larger, IP66-rated weatherproof housing, PoE-only connectivity, and a mounting bracket. While both cameras share the same IMX500 sensor and Raspberry Pi CM5 platform, they are clearly differentiated by form factor, environmental protection, and intended use cases.
Another camera based on the Sony IMX500 AI camera sensor is the LUCID SENSAiZ Smart camera, but it’s closer to the design of the B0581 model, and is not designed for outdoor use like the B0572 SKU.
The Arducam All-in-One Raspberry Pi AI Camera Kit with CM5 (B0572) is listed for $379.99. The package includes the camera unit (with CM5 pre-installed) and a 1-meter Ethernet cable. The more compact indoor version (B0581 SKU) also sells for $379.99. Both models are also listed on AliExpress, but for crazy prices ($700+), so we don’t recommend purchasing from there at this time. Another (future) option may be the company’s Amazon store, but neither product is listed at the time of writing.
Debashis Das is a technical content writer and embedded engineer with over five years of experience in the industry. With expertise in Embedded C, PCB Design, and SEO optimization, he effectively blends difficult technical topics with clear communication
Support CNX Software! Donate via cryptocurrencies, become a Patron on Patreon, or purchase goods on Amazon or Aliexpress. We also use affiliate links in articles to earn commissions if you make a purchase after clicking on those links.
