reComputer AI Industrial R
Raspberry Pi CM5 powered industrial edge AI system with Hailo-8 26 TOPS acceleration
Introduction
The reComputer AI Industrial R2135 is powered by Raspberry Pi CM5 and a Hailo AI accelerator. This compact edge AI system delivers 26 TOPS for real-time multi-channel vision processing. With a quad-core Cortex-A76 CPU, up to 16GB RAM, up to 64GB eMMC, and a versatile interface set, it is designed for seamless integration into industrial AI applications.
Features
- Powerful performance with Raspberry Pi CM5 module: 2.4GHz quad-core 64-bit Arm Cortex-A76 and VideoCore VII GPU, with up to 16GB RAM.
- AI-optimized computing: Hailo-8 AI accelerator delivers up to 26 TOPS for real-time, high-performance AI vision processing across multiple channels.
- Rich connectivity: 2 x HDMI, 1 x Gigabit Ethernet, 2 x USB 3.2 Type-A, 1 x USB 2.0 Type-C, plus slots for SSD storage and AI acceleration.
- Advanced video processing: 4K 60fps H.265 decoder supporting 2-8 real-time 720p multi-stream inferences for object detection and tracking.
- Enhanced industrial reliability: includes additional PoE, 4G/LoRa, TPM, and RTC options.
- Customization supported: contact edge@seeed.cc for customization needs.
Hardware Overview
Applications
With advanced features such as object detection, object recognition, object tracking, OCR, and abnormal event detection with real-time alert triggers, this AI box is designed to deliver strong performance in resource-constrained environments.
Surveillance & Security
Integrated into NVR systems for real-time video analytics, including object detection such as intruders or abandoned items, plus anomaly detection to enhance situational awareness and security efficiency.
Smart Retail
Create smart retail solutions using AI-driven features such as real-time visual product recognition, personalized shopping experiences, and advanced customer behavior analytics for optimized store operations.
Smart Transportation
Process traffic and parking data from cameras and sensors at the edge, enabling intelligent traffic flow analysis and smart parking guidance to reduce congestion.
Digital Signage & Audience Engagement
Connect cameras and content servers for dynamic content adaptation based on facial recognition and audience demographics, improving engagement in public spaces and retail environments.
Smart Building
Integrate sensors and IoT devices to monitor and optimize lighting, HVAC, security, energy efficiency, predictive maintenance, and occupant comfort.
Part List
| Item | Quantity |
|---|---|
| reComputer AI Industrial R2000 | 1 |
| DIN-rail clip | 1 |
| Screw Pack | 1 |
| DC Power Female Jack to Screw Terminal Adapter | 1 |
| 12V/3A Power Adapter (with 4 interchangeable adapter plugs) | 1 |
| User Manual | 1 |
Product Selection
| Product Model | SKU | Memory | Storage | AI Accelerator |
|---|---|---|---|---|
| reComputer AI Industrial R2135-12 | 114993595 | 8GB RAM | 32GB eMMC | 26 TOPS |
| reComputer AI Industrial R2145-12 | 103990763 | 16GB RAM | 32GB eMMC | 26 TOPS |
| Parameter | reComputer AI Industrial R2000 |
|---|---|
| Product Series | reComputer AI Industrial R2000 |
| Platform | Raspberry Pi Compute Module 5 |
| CPU | 2.4GHz quad-core 64-bit Arm Cortex-A76 |
| GPU | VideoCore VII |
| AI Processor | Hailo-8 M.2 Acceleration Module, 26 Tera-Operations Per Second |
| RAM | 2GB/4GB/8GB/16GB |
| eMMC | 0GB/16GB/32GB/64GB |
| Operating System | Raspberry Pi OS, Ubuntu |
| Video Decoder | 4Kp60 HEVC decoder |
| Power Supply | DC 9V-36V, 2-pin Terminal Block |
| USB | 2 x USB 3.2 USB-A; 1 x USB 2.0 USB-C for debug and update |
| Display | 2 x Standard HDMI Ports, HDMI 2.0 |
| Ethernet | 1 x 10/100/1000 Mbps RJ45 |
| M.2 Slot | 1 x USB 3.0 to M.2 slot, M.2 M-key 2280; 1 x PCIe 3.0 to M.2 slot, key 2242 |
| Mini-PCIe Slot | 1 x Mini-PCIe for 4G/LoRaWan module |
| SIM Card Slot | 1 x Standard SIM Card Slot |
| Audio | 1 x 3.5mm Audio Jack |
| Wi-Fi | Dual-band 802.11ac Wi-Fi |
| Bluetooth | Bluetooth 5.0 / BLE |
| 4G Cellular | 4G LTE (Optional) |
| Ingress Protection | IP40 |
| Operating Temperature | -20~65 C |
| Warranty | 2 Years |
| Watchdog | Hardware WatchDog |
| RTC | High Accuracy RTC |
| Security | Encryption Chip TPM2.0/ATECC608A (Optional) |
| Heat Dissipation | Heatsink with Fan |
Documentation
- reComputer User Manual
- Seeed reComputer AI R Series Flyer
- Seeed Rasperry Pi Flyer
- Seeed reComputer AI R Series Flyer-CN
- Seeed Raspberry Pi Flyer-CN
From Pixels to Meaning: How Does Raspberry Pi CM5 AI Box See and Interpret?
Deep dive into how a Raspberry Pi CM5 AI Box turns raw pixels into real-time understanding of objects, movements, and scenes, including fall recognition, fence climbing, and CLIP demos.
Moonfire NVR on reComputer
Moonfire-NVR is a lightweight NVR system designed to record from IP cameras directly to a file system or cloud storage.
go2rtc on reComputer
go2rtc is an open-source WebRTC framework combining WebRTC with RTSP, RTMP, and HLS for efficient real-time video and audio transmission.
Seeed launched a Raspberry Pi CM5/Pi 5 AI Box with Integrated Hailo NPU
The reComputer AI R Series combines Raspberry Pi CM5/Pi 5 with the Hailo-8 NPU for powerful edge AI capabilities.
Master AI with Raspberry Pi, Power Smart Retail Solutions
A roadmap covering quick-start Raspberry Pi configuration, hands-on AI framework learning, and real-world model applications.
How Does the World Most Powerful Raspberry Pi CM5 AI Box See and Understand?
Overview of the reComputer AI Industrial R as a Raspberry Pi AI Box for vision AI systems.
Making Next Gadget - See It Live: Raspberry Pi CM5 AI Box Launch + Hands-On Tutorial Updates
Hands-on content including hardware teardown for SSD, 4G, LoRa, AI module, Python examples, WebUI hacks, and demos with Home Assistant and Frigate.
Getting Start with reComputer AI Industrial R2135 Series
Getting started with the reComputer AI Industrial R2135, powered by Raspberry Pi CM5 and Hailo AI accelerator for 26 TOPS real-time multi-channel vision processing.