Put intelligence closer to the work
The Converged Edge Center of Excellence explores the meeting point of computing, connectivity, data and operational technology. The focus is the practical work of building and managing intelligence near devices, people and physical processes.
Overview
The Converged Edge CoE helps practitioners design and operate edge platforms for real workloads across industrial, commercial and remote sites. Its reference architectures, playbooks, evaluations and evidence address data, control and AI near the source, including offline operation, security and cost.
Chair: Cole Edmundson (Oracle).
Audience & collaboration
For edge and network teams, site operators, data/AI developers, security and risk leaders, executives, OEMs, integrators, telecom providers and academia.
How we work
Share deployment patterns and roadmaps, compare approaches across organizations, and publish concise guidance through virtual and live sessions.
Participation
Members learn and review; contributors share use cases and artifacts; workstream leads deliver outputs; chair and steering members set direction and quality standards.
Charter & objectives
Build edge systems that remain reliable offline, recover quickly, provision and update safely, and scale across sites with limited local support. Right-size compute and connectivity for latency and cost; use open interfaces, identity, policy gates and audit trails.
Pillars & architecture
- Platforms: lightweight K8s, containers, WASM, signed images and staged rollback.
- Networks: wired, Wi-Fi and private 5G; VLAN/VRF, TSN where needed, QoS and isolation.
- Data: local streams and stores, retention, buffering, compression and controlled cloud sync.
- Resilience: redundancy, UPS awareness, brownout behavior, partition tolerance and recovery drills.
- Security: identity, attestation, SBOM/VEX, sealed secrets and secure updates.
- Operations: metrics, logs, traces, fleet health, SLOs and alert-linked runbooks.
- Workloads: CPU/GPU/NPU/DSP sizing for vision, time series and real-time tasks, including energy and thermal limits.
Reference architecture
Field devices and gateways connect through segmented networks to signed edge workloads, local data services and enterprise applications. Governance adds RBAC/ABAC, human approvals for sensitive actions, rollback and audit evidence.
Outputs & cadence
Outputs: reference architectures with design rules and BOM considerations; evidence packs linking policy, controls, logs and approvals; and Masterclasses on orchestration, data, networking, resilience and governance.
Cadence
Quarterly showcases (2–3 hours) and annual IoT Slam deep dives and Masterclasses.
Get involved
Join the CECoE and choose a workstream at Join Us. To sponsor a workstream or propose a pilot, email sponsors [at] iotcommunity [dot] net.
Bring an architecture decision or deployment lesson to help others weigh performance, resilience, cost and control.
A question worth exploring
Which decisions belong at the edge, and how will the resulting system be operated, secured and supported across multiple sites?
Bring your perspective
Bring an architecture decision, a deployment constraint or a lesson from operating at the edge. Help others assess the trade-offs between performance, resilience, cost and control.