As utilities define their AMI 2.0 and Grid Edge Intelligence roadmaps, the conversation increasingly starts with the use cases. Not just meter-to-cash, but the next horizon: advanced transformer load monitoring, enhanced outage insights with waveform-based fault visibility, voltage and PQ analytics, EV charging coordination and additional sophisticated behind-the-meter interactions. 

These use cases behave very differently from traditional AMI 1.0 workloads. They are more mission-critical, more distributed, requiring more frequent yet reliable communication, and more tightly coupled to distribution system operations. In many ways, they sit closer to SCADA and Distribution Automation in terms of timing expectations and operational criticality. 

As utilities expand into AMI 2.0 and Grid Edge Intelligence, the challenge is no longer simply selecting a communications technology but determining how to evaluate the right network approach for their specific needs. What we present here is a practical process: begin by identifying your priority use cases, understand the traffic and timing characteristics they introduce and then allow those requirements to guide which network architecture and capabilities best support them. 

STEP 1: Recognize What AMI 2.0 Really Demands 
When mapping out AMI 2.0 use cases, two themes consistently surface: 

1. Many new applications require predictable, near-real-time communication. Minute-level transformer loads, sub-minute PQ exceptions, waveform signatures associated with emerging grid faults and high-frequency DI applications. 
2. Traffic grows in two directions: local cluster traffic between meters under the same transformer, and back office-bound traffic delivering analytics data and events upstream. 

These two flows both increase materially in AMI 2.0 and lead to the question: What should be the baseline communications foundation for AMI 2.0? 

STEP 2: Evaluate Private LTE as a Natural Foundation 
Private LTE offers operational control, improved lifecycle alignment, independence from public carrier timelines and a unified platform for field operations and automation. It is logical to explore how it can also serve as the baseline for AMI 2.0. 

STEP 3: Run the Models and Discover Constraints 
Traffic modeling reveals three consistent findings: 

1. Achieving near 100% coverage is feasible but can be expensive, as AMI requires reach into meter rooms, basements, underground service, rural terrain and other hard‑to‑reach environments. 
2. When every meter sends frequent small data transmissions directly on the cellular network, the cellular layer becomes burdened with simultaneous device activity and signaling overhead - something wide‑area networks are not optimized for at AMI scale. 
3. Even in private LTE, operating and managing many small “pipes” typically carries higher cost and complexity than supporting fewer, higher‑capacity ones — just as carriers experience. 

STEP 4: Consider Architectural Patterns That Solve Identified Challenges 
Two practical architectural patterns have emerged that support AMI 2.0 more efficiently. 

Approach A: Micro Mesh with Higher-Category Backhaul 
A small cluster, low-hop RF mesh network deployed with a flexible set of takeout point options that provide WAN backhaul and utilize a higher-category cellular modem to deliver traffic to the back-office. This can be augmented with PLC where appropriate that offloads local transformer-cluster traffic. Benefits include predictable latency, aggregated traffic and a cellular LTE/5G category, improved reach via mesh relaying and the ability to separate long-life meters from shorter-lifecycle backhaul equipment for rate-case flexibility. 

Approach B: Cellular + Grid-Edge P2P w 
Every meter includes an LTE Cat-M1 modem, but not all individual meters transmit frequent small data transmissions to the WAN. Instead, they coordinate locally and elect a spokesmeter at the application level (within the DI environment) to aggregate and forward back office-bound data. Benefits: reduced WAN signaling, structured and predictable upstream traffic, directly connected paths available when needed (bellwether outage notifications, urgent alarms), and coexistence of Cat-M1 and future eRedCap devices on a unified private LTE/5G network. 

STEP 5: Validate Network Design for Scalability and Future Growth  
What we have seen is that both approaches align well with the Itron’s Intelligent Connectivity pillars that have emerged from learnings of many AMI utility deployments: 

• Enduring Longevity: Meter endpoints maintain up to 20-year life, while takeout points evolve separately and more frequently. 
• Resilient Reliability: Multi-path redundancy, self-optimizing network behavior and coordinated pacing and traffic prioritization via a centralized traffic manager. 
• Security by Design: Zero Trust principles and end-to-end encryption at application and network level with secure enrollment and renewal processes that reinforce device and application identity throughout the lifecycle. 
• Expansive Coverage: Sub-GHz RF & cellular for good propagation, relay and satellite extensions for hard-to-reach meters. 
• Flexibility & Scalability: Flexible set of AP variants, transformer-level spokes meter aggregation, seed-and-expand firmware distribution. 

A Natural Conclusion: A Balanced AMI 2.0 Network 
The approach outlined above reflects a natural process: start with the use cases, understand their requirements, evaluate fit with existing and planned unified networks, identify constraints, introduce structures that solve them and validate network design for scalability and future growth. From this set of steps, a balanced AMI 2.0 network emerges, leveraging the strengths of each technology where it fits best. 

Stop by the Itron booth (#2715) at DTECH in San Diego, CA from Feb. 2-5 for more information about AMI 2.0 and Grid Edge Intelligence roadmaps, or feel free to send an email to jan.forslow@itron.com.