Scaling the Industrial Internet of Things (IIoT) makes sending all raw data to the cloud economically and technically inefficient. Modern architecture is built on a pragmatic workload distribution between edge computing and cloud platforms.
Workload distribution criteria
To determine the optimal data processing level, architects must evaluate the system based on several key parameters:
- Latency: If response time is critical for local safety loops (less than 10 ms), computing must occur exclusively at the edge level.
- Autonomy: According to the NIST SP 800-82 guide, system availability is a priority. Local control loops must function even in the event of lost cloud connectivity.
- Traffic optimization: Edge devices allow filtering high-frequency telemetry and normalizing data using the OPC UA standard before transmission.
The role of the cloud and infrastructure security
The cloud level remains essential for tasks requiring long-term data storage, global coordination, and training predictive maintenance models. Optimized analytical models are subsequently deployed to the edge for rapid local analysis.
Integrating industrial networks with the cloud requires strict compliance with security standards. The ISA/IEC 62443 series mandates clear segmentation of IT/OT networks to prevent unauthorized access to industrial equipment.
Practical implementation of hybrid architectures
Designing such systems is a custom engineering task. Expertise in building hybrid architectures and implementing AI algorithms on edge devices in Ukraine is offered by Softengi (a member of Intecracy Group). To build the upper level of ecosystems—enterprise monitoring portals and data aggregation from thousands of edge nodes—the domestic low-code platform UnityBase is used.
Market implications
Failing to balance edge and cloud computing leads to high data transmission costs and operational risks due to latency. Adopting a hybrid architecture allows industrial businesses to maintain continuous operations during network outages, reduce cloud storage costs through local filtering, and protect critical infrastructure from cyber threats.
Recommendations
- Analyze system requirements: Evaluate latency, autonomy, and traffic optimization needs to determine which workloads belong at the edge versus the cloud.
- Enforce security standards: Segment IT and OT networks strictly according to ISA/IEC 62443 to prevent unauthorized access.
- Deploy hybrid solutions: Partner with experienced integrators like Softengi to implement edge AI, and utilize platforms like UnityBase for high-level enterprise data aggregation.
Prepared by a Software Ukraine member. Original publication.