Edge Runtime

Guarantees that control logic completes within a defined time budget, measured in microseconds, not milliseconds. This is non-negotiable for closed-loop motion and safety.

• PREEMPT_RT Kernel: Applies kernel patches to Linux, making it fully preemptible for hard real-time scheduling.
• CPU Pinning: Dedicate specific physical cores exclusively to the runtime, eliminating cache misses and scheduler jitter.
• Time-Sensitive Networking (TSN): Leverages IEEE 802.1 standards for deterministic, low-latency data delivery over converged Ethernet networks.
• Precision Time Protocol (PTP): Synchronizes distributed clocks to sub-microsecond accuracy via IEEE 1588, essential for coordinating isochronous cycles across nodes.

Decouples software-defined control logic from proprietary physical controllers, enabling workload portability and consolidation.

• Type-1 Hypervisor: A bare-metal hypervisor runs directly on silicon, providing the highest determinism and resource isolation for mixed-criticality workloads.
• Single Root I/O Virtualization (SR-IOV): Allows a single physical NIC to present as multiple virtual devices, giving VMs direct I/O access without hypervisor overhead.
• Soft PLC: Executes standard IEC 61131-3 control logic on general-purpose compute, replacing proprietary hardware controllers.
• Immutable Infrastructure: Deploys control components as pre-configured golden images, replaced entirely rather than patched, ensuring absolute configuration consistency.

The runtime must function independently of cloud connectivity, executing control logic, protocol translation, and local analytics even during WAN outages.

• Local Decision Loop: All critical path logic executes locally; the runtime never depends on a round-trip to a remote server for actuation.
• Edge-Native Data Store: A local time-series database buffers telemetry during disconnection, forwarding data upon reconnection.
• MQTT Sparkplug: A mission-critical specification for MQTT that adds standardized topic namespaces, payload definitions, and state management, enabling efficient local broker architectures.
• Unified Namespace (UNS): A local semantic data hierarchy aggregating all data sources, allowing any application to discover and consume real-time information via a common interface.

Runs safety-critical control, non-critical HMIs, and analytics on a single edge server with strict temporal and spatial isolation.

• Safety Integrity Level (SIL): Discrete risk reduction measures defined by IEC 61508, dictating rigorous development requirements for safety functions.
• Fault Tolerance (FT): A secondary redundant system executes in lockstep, enabling instantaneous, bumpless takeover without state loss upon hardware failure.
• Workload Consolidation: Merges discrete control, HMI, and analytics onto one high-performance server, reducing hardware footprint, cabling, and energy consumption.
• Real-Time Hypervisor: Hosts both RTOS and GPOS on shared silicon without compromising microsecond-level latency for critical tasks.

Packages control functions, protocol translators, and analytics as lightweight, portable containers managed by declarative infrastructure.

• Infrastructure as Code (IaC): Manages and provisions industrial control infrastructure through machine-readable definition files, enabling version-controlled, repeatable deployments.
• Live Migration: Moves a running virtualized control workload between physical hosts without interrupting execution state, enabling zero-downtime maintenance.
• Virtual Commissioning: Validates and debugs PLC code and HMI interfaces against a digital twin of the production cell before physical installation, drastically reducing on-site startup time.
• IEC 61499: Defines a component-based function block architecture for distributed automation, enabling event-driven control logic decoupled from specific hardware topologies.

Bridges the gap between legacy industrial protocols and modern IT systems by translating and contextualizing data at the edge.

• OPC UA Pub/Sub: Enables scalable, connectionless data distribution using a publish-subscribe pattern, often combined with TSN for deterministic field-level communication.
• Data Distribution Service (DDS): A data-centric publish-subscribe middleware standard for real-time, decentralized architectures common in autonomous systems.
• Data Processing Unit (DPU): A specialized programmable accelerator that offloads networking, security, and storage virtualization from the host CPU, freeing cycles for real-time control.
• Virtualized Network Function (VNF): A software instance of a network service, such as a protocol router, running on a hypervisor instead of a dedicated appliance, enabling dynamic service chaining.