All resources
Whitepaper · 15 Sept 2026 · 16 min read · smoc.cloud team

Security operations for AMI: monitoring keys, events and tenants

How utilities can monitor security events, key lifecycles and access across a multi-vendor AMI estate and produce audit evidence on demand.

Executive summary

Smart metering estates are now part of critical infrastructure. They hold customer data, carry commands that can disconnect supply or close a gas valve, and depend on millions of cryptographic keys. Yet in most utilities, security monitoring for metering is split across head-end systems that each log events differently.

This paper sets out how to run security operations for Advanced Metering Infrastructure (AMI) as one discipline: one event stream, one view of key health, strong tenant isolation and an audit trail that regulators trust.

Figure 1. Security events from several head-end systems normalised into one security operations view.
Figure 1. Security events from several head-end systems normalised into one security operations view.

1. Why AMI security is hard to see

  • Different event formats: each vendor names, scores and timestamps security events in its own way
  • Split key management: keys are generated, stored and rotated in different systems with no shared record
  • High-impact commands: disconnects, valve closures and mass firmware updates can be issued from several consoles
  • Evidence on request: audit evidence is assembled by hand when a regulator or customer asks
Time to answer a typical audit question
Who sent this command?Which keys are overdue?Failed auth by regionFirmware integrity0h4h8h12h16h
  • Fragmented tooling
  • Unified security operations
Figure 2. Illustrative. Questions that take days across vendor consoles become a filtered report.

2. One normalised event stream

The first step is to bring every security-relevant event into one stream with common definitions: authentication failures, invalid frame counters, tamper alarms, unexpected configuration changes, firmware signature failures and command executions.

Normalisation makes correlation possible. A burst of authentication failures on meters from two vendors behind the same concentrator is one incident, not two.

Security events after normalisation and correlation (per day)
Raw eventsDe-duplicatedCorrelatedIncidents030006000900012000
Figure 3. Illustrative. Correlation turns thousands of raw events into a handful of incidents an analyst can act on.

3. Key lifecycle monitoring

Every DLMS/COSEM association depends on keys. Losing track of them is one of the most common gaps in AMI security.

Figure 4. The key lifecycle: generation, distribution, use, rotation and revocation.
Figure 4. The key lifecycle: generation, distribution, use, rotation and revocation.

Track for each meter and association:

  1. Generation and source of each key
  2. Distribution and confirmation that the meter accepted it
  3. Usage, including which operator or system used it and when
  4. Rotation against policy, with overdue keys flagged automatically
  5. Revocation when a meter is removed, replaced or suspected compromised
Share of meters with keys rotated within policy
M0M3M6M9M120%25%50%75%100%
Figure 5. Illustrative. Visibility alone drives rotation compliance up, because overdue keys stop being invisible.

4. Tenant isolation and least privilege

Where one platform serves several utilities, business units or client estates, isolation must be structural, not just a filter in a report.

Figure 6. Each tenant's meters, data and keys are isolated on a shared, secured platform.
Figure 6. Each tenant's meters, data and keys are isolated on a shared, secured platform.
  • Separate data and keys per tenant, with no shared access paths
  • Role-based access so analysts, field coordinators and administrators see only what they need
  • Four-eyes approval for disconnects, valve operations and bulk firmware updates
  • Time-limited elevated access, logged and reviewed

5. An audit trail regulators trust

Every command, configuration change, key operation and access decision is written to an immutable log with who, what, when and why. Reports for regulators, auditors and customers then become a query rather than a project.

6. A staged path

  1. Months 0 to 3: connect security events from every head-end and publish one risk view
  2. Months 3 to 6: add key lifecycle tracking and overdue-key alerts
  3. Months 6 to 12: enforce approvals for high-impact commands and automate audit reporting

Conclusion

AMI security does not need another tool per vendor. It needs one operating view of events, keys, tenants and commands. Read more about our security approach and security architecture, or request a demo.

Figures marked illustrative are indicative examples and not results from a specific customer.

How smoc.cloud helps

Security operations across a multi-vendor AMI estate means watching keys, events and access in one place. smoc.cloud gives the security team that visibility without stitching vendor logs together.

  1. Spot it

    Monitor keys, events and tenants

    Key lifecycles, security events and access activity across all head-ends are monitored centrally, with alerts for expiring keys and unusual behaviour.

  2. Find the root cause

    Correlate to find the real threat

    Events are linked across meters, networks and users, so a coordinated pattern stands out from routine noise and the affected scope is clear.

  3. Fix it

    Respond with evidence

    Containment actions, key rotation and access changes are tracked in an immutable audit trail, producing compliance evidence on demand.

What this means for you

Critical national infrastructure obligations mean security events need to be visible across every vendor. One normalised stream makes NIS-style reporting a report, not a project.

DNOs: how smoc.cloud helps
The Metering Operations Brief

One email a month with new articles, whitepapers and field notes on running smart metering estates. Unsubscribe at any time.

Join the conversation on LinkedIn

Open on LinkedIn
Nov 18, 2026

Customers never see your head-end. They see their bill. A commissioning that stalls. A meter that quietly stops reading. A register that drifts. None of it reaches the customer straight away. Weeks later it arrives as an estimated bill and a call to the contact centre. Our new whitepaper follows five moments that shape the customer experience, from installation to support, and shows where operations decide whether each one goes right. Which stage generates most of your metering complaints? #CustomerExperience #SmartMetering #Utilities

Nov 18, 2026

Two meters. Both fully DLMS/COSEM compliant. Different object lists, different event codes for the same power failure, different load profile settings. Compliance gives you a common language. It doesn't give you a common vocabulary. That's the gap the DLMS UA Generic Companion Profile closes: standard objects and behaviours that every model implements the same way. Fewer custom integrations, faster onboarding of new models, less lock-in. Our new whitepaper explains what it means for the metering business, not just the engineers. #DLMS #SmartMetering #Standards

Nov 17, 2026

If you're a utility lead, your IT and compliance teams will ask about security before you ever get to features. They should. Metering platforms touch critical infrastructure and customer data. A good demo answers the hard questions early: how the platform is secured, how it handles compliance, how it fits protocols you already run. What's the first question your IT team asks about a new platform? #UtilityOperations #SmartMetering #CyberSecurity

next_step

See your estate in smoc.cloud.

A demo walks through your utility types, your vendors and your compliance obligations — not a generic slide deck.