Manchester Edge AI Lab

Manchester Edge AI Lab is where Resolutis’s embedded systems expertise meets the manufacturing sector. We deliver edge AI integration, inference observability, predictive maintenance, and compliance services, built for manufacturers, retailers and supply chain who need production-grade edge AI, not cloud-dependent workarounds.

Unlike cloud-first platforms that stop short of the device, our work starts at the edge, where latency is measured in microseconds and uptime is non-negotiable.

The Problem We Solve

Most Northern manufacturers face three compounding edge AI challenges:

• Inference runs blind — models deployed to factory/shop floor hardware have no visibility into latency, drift, or failure. Standard monitoring tools don’t reach the device.
• Cloud tools don’t fit — Datadog, Grafana, and Azure Monitor were not built for Jetson Orin or NXP i.MX 9x. Adapting them to constrained hardware is expensive and unreliable.
• Embedded AI talent is scarce — hiring engineers who understand both embedded software and production AI can be difficult.

Services

01 — EdgeOps: Inference Observability

Real-time visibility into your edge AI runtime. EdgeOps monitors inference latency, GPU utilisation, model throughput, and thermal headroom with a focus on NVIDIA Jetson and NXP i.MX deployments.

Purpose-built for constrained devices. No Kubernetes required. Rust-native agent, sub-1% CPU overhead, designed for 24/7 factory conditions. Think Datadog for edge inference — without the cloud dependency.

• Inference latency and throughput monitoring
• GPU and NPU utilisation tracking
• Thermal and power monitoring
• Model drift alerting
• REST API and dashboard
• Supports NVIDIA Jetson Orin and NXP i.MX 9x

02 — Edge AI Integration Services

Getting a model running on a development board is the easy part. Getting it running reliably on a factory floor — integrated with your existing systems, monitored, and maintainable is where most projects stall. That’s what we do.

• Edge-to-ERP data pipeline design and integration
• OPC-UA, Modbus, and MQTT connectivity
• Model deployment and lifecycle management
• Predictive maintenance applications
• Compliance and audit logging for industrial deployments
• On-premises, no cloud dependency required

03 — Embedded AI Consulting

Fractional CTO and embedded systems engineering for manufacturers deploying edge AI in production. From device driver development and wireless connectivity to full edge inference pipeline design.

• Fractional CTO — embedded AI leadership
• Wireless connectivity — LTE-M, NB-IoT, BLE
• Zephyr RTOS and Rust systems programming
• Linux BSP
• Edge inference pipeline design
• NXP and NVIDIA ecosystem expertise