Rugged Edge AI Computers & Fanless Industrial PCs
Compare your platform Talk to an engineer
Choose an edge AI computer based on its deployment environment and processing requirements. Indoor and in-vehicle applications can use dedicated edge AI platforms, while exposed installations require IP66-rated outdoor models. For camera-pole deployments that also require integrated camera power and networking, Vision Guard combines these functions in one system.
Which edge AI computer is right for your application?
EDGE AI PLATFORMS
Rugged Edge AI Computers for Local Inference
Fanless Mini METIS
FM-METIS-51 · Axelera Metis
Fanless edge AI computing powered by the Axelera Metis AIPU, delivering up to 214 TOPS of AI acceleration alongside an eight-core Arm processor and 16GB memory. Designed for indoor machine vision and analytics, this compact IP51 platform operates within a 9–20 W power envelope.
View ProductHybrid Cooled Mini Jetson Orin 32GB
HCM-JON-32-41 · Jetson Orin
Hybrid-cooled edge AI computing powered by NVIDIA Jetson Orin, featuring a 1,792-core Ampere GPU, 32GB memory and 10GbE networking. Designed for GPU-accelerated video analytics, with CAN bus and serial interface options in a compact IP41 chassis.
View ProductHybrid Cooled Mini Jetson Orin 64GB
HCM-JON-64-41 · Jetson Orin
Hybrid-cooled edge AI computing powered by NVIDIA Jetson Orin, featuring a 2,048-core Ampere GPU, 64GB memory and 10GbE networking. Designed for demanding video analytics and concurrent AI workloads, this compact IP41 platform supports applications running multiple models in parallel.
View ProductFanless Mini METIS In-Vehicle
FM-METISIV-50-12V · Axelera Metis · In-vehicle
Fanless in-vehicle edge AI computing powered by the Axelera Metis AIPU, delivering up to 214 TOPS of AI acceleration. Featuring intelligent ignition control and a 12 V vehicle power supply, this compact IP50 platform is designed for onboard AI inference and analytics in mobile applications.
View ProductOUTDOOR EDGE AI COMPUTERS
Outdoor edge AI computers sealed to IP66
Fanless Mini METIS IP66
FM-METIS-66 · Axelera Metis
High-performance, fanless outdoor edge AI computing powered by the Axelera Metis AIPU, delivering up to 214 TOPS of AI acceleration. With dual Gigabit Ethernet and a sealed, IP66-rated waterproof chassis, the platform supports exposed camera installations, roadside infrastructure and outdoor machine vision.
Choose the power configuration to suit your installation:
FM-METIS-66 · AC powered: 110–220 V mains input for outdoor AI deployments with an available AC supply.
FM-METIS-66-12V · DC powered: 12 V DC input for roadside and remote camera installations where low-voltage power is required.
View ProductINTEGRATED AI COMPUTER
An integrated AI computer for camera poles
Vision Guard
VG-15MV-66 · Pole-mounted
Fanless edge computing powered by an Intel Core Ultra 5 225H processor, with three PoE camera ports (90 W in total) and an optional Axelera Metis accelerator at up to 214 TOPS. Built for pole-mounted camera positions from −20°C to 70°C, in an IP66 waterproof enclosure.
View ProductPRODUCT SELECTION
Compare Compact Edge Systems by Key Specifications
Compare the three Tranquil IT compact edge computing platforms across key technical specifications. Full product details are available on each individual product page.
WHY ON-DEVICE AI
The benefits of running AI inference locally
Reduced network traffic
Processing video and sensor data at the edge reduces the volume of information sent across the network, helping conserve bandwidth.
Local data processing
Images and sensor data can be processed on-site, helping keep sensitive information within the local environment, subject to system architecture
Inference without continuous connectivity
On-device AI inference can continue when network connectivity is limited or unavailable, provided the application and required data are available locally.
Power-efficient AI acceleration
Axelera Metis platforms deliver AI inference acceleration within a defined power envelope, supporting deployments where energy consumption and thermal constraints matter.
Logistics, Ports & Air Cargo
Edge AI for visual inspection, cargo monitoring and operational analytics across sorting facilities, access gates, freight yards, cranes and yard vehicles.
In-Vehicle
In-vehicle edge AI computing for onboard video analytics and real-time processing, powered directly from the vehicle’s electrical system.
RECOMMENDED FOR
Compact Edge AI Systems for Industrial and Commercial Applications
From healthcare workstations and surveillance systems to ANPR and digital signage, compact edge computers support a range of dedicated applications. Explore the recommended Tranquil IT platforms for each use case, with model selection guided by connectivity, display requirements, mounting options and deployment conditions.
Surveillance
AI video analytics for indoor monitoring, industrial sites, outdoor camera installations and pole-mounted surveillance systems.
Healthcare
Local AI inference alongside clinical equipment, supporting application-specific image analysis and monitoring within the relevant clinical and regulatory requirements.
ANPR
Edge AI for automatic number plate recognition at roadside cameras, including exposed installations where IP66 protection and low-voltage DC power are required.
AI CONFIGURATION & CUSTOMISATION
Deploy Your Models, or Adapt the Hardware
AI deployments may require software configuration, model preloading or changes to the computing hardware. The right approach depends on whether the existing platform meets the physical and technical requirements of the application.
Tranquil Services
AI models, containers, applications and site-specific settings can be preloaded onto a standard platform. Tranquil Services supports software configuration and integration with compatible partner applications, without changing the physical hardware.
Custom AI Hardware and System Integration
When an application requires a non-standard form factor, custom interfaces or an integrated camera and sensor system, Tranquil Studio assesses the engineering requirements. Solutions may include adapting an existing platform or developing purpose-built AI hardware.
Recommended FAQ section
An edge AI computer runs trained AI models locally, near the camera, machine or vehicle generating the data. This enables AI inference close to the source, reducing reliance on remote processing and network connectivity.
Axelera Metis platforms pair a dedicated AI Processing Unit (214 TOPS) with an eight-core Arm processor, for supported vision and inference workloads. NVIDIA Jetson Orin platforms combine an Ampere GPU, Tensor Cores and an Arm CPU, which suits models developed on NVIDIA tools. The right choice depends on the AI workload, software framework and processing requirements.
The FM-METIS-51, FM-METISIV-50-12V, FM-METIS-66, FM-METIS-66-12V and VG-15MV-66 use fanless, passive cooling. The HCM-JON-32-41 and HCM-JON-64-41 Jetson Orin models use hybrid cooling.
Axelera Metis platforms use the Voyager SDK, which supports model compilation from PyTorch and ONNX, alongside a model zoo covering object detection, image classification, segmentation and small language models. NVIDIA Jetson Orin platforms support AI workloads developed using the NVIDIA software ecosystem. Tranquil Services can assist with model deployment, software integration and pre-configuration before dispatch, subject to platform compatibility and application requirements.
The FM-METIS-66 and FM-METIS-66-12V support operating temperatures from 0°C to 60°C. The Vision Guard (VG-15MV-66) operates from −20°C to 70°C, making it suitable for a wider temperature range in pole-mounted deployments. Vision Guard also integrates camera power and networking to simplify outdoor installation.
Yes. The FM-METISIV-50-12V is designed for in-vehicle use, with a 12 V supply and intelligent ignition control. The FM-METIS-51 and FM-METIS-66-12V run from 12 V DC, and the two Jetson Orin models accept 9–36 V DC. Only the FM-METISIV-50-12V is designed for vehicle installation.
Yes. Tranquil Services can preload compatible software, AI models and configuration settings without changing the physical platform. Where a requirement involves hardware adaptation or a bespoke design, Tranquil Studio assesses the engineering requirements.
Tell us where your edge AI has to run.
Cameras or sensors, model and workload, environment, power, connectivity and number of sites. An engineer will recommend a platform, or a validated solution with partner software.