Harnessing the power of pLTE and AWS Cloud to optimize AMI 2.0 outcomes

AMI 2.0 has gained traction since the early 2020s, promising real-time insights, enhanced grid management, and demand response. By 2029, U.S. smart meter installations will reach 182.9 million, covering 94% of customers. However, AMI 2.0 isn’t just about meters, it demands high-bandwidth communication and scalable storage. A single meter now generates 1,440 data points daily, driving terabytes of data.

Without robust infrastructure, benefits such as real-time outage detection and predictive maintenance remain unrealized. Utilities must not only adopt smart meters but also build the digital backbone that supports them for full AMI 2.0 potential.

What can I get out of an upgrade to AMI 2.0?

Currently, most utilities across the United States are operating AMI 1.0. For some utilities, this might even only be a very rudimentary ”AMI 0.5” with minimal automation, delayed data collection, and little to no real-time communication capabilities. Legacy systems won’t be polling anywhere near as regularly when compared to what would occur following an upgrade to AMI 2.0. Why that matters is that the many touted benefits of AMI 2.0 can only be achieved with far more frequent polling. The types of gains you could expect include:

• Bi-directional communication: this upgrade alone offers an enhanced ability to send commands to meters and to gain more control over load management.
• Enhanced security: offering better encryption and more robust cybersecurity protocols, as well as safeguards of sensitive information, AMI 2.0 provides greater security than legacy metering systems.
• Data analytics: with vast amounts of data available for crunching, one of the most transformative benefits of AMI 2.0 centers on analytics. From consumption patterns and peak usage through to predictive maintenance, data analytics become a key tool for utilities to assist with aspects such as fault detection and load management.
• Consumer control and empowerment: offering granular usage data, AMI 2.0 is also transformational at the end user level by enabling the ability to modify energy usage as desired to influence billing.
• Ability to remotely monitor: increasing polling rates significantly and with the inclusion of bi-directional communication allows utilities to remotely monitor (and manage) meters, which can help with the detection of outages or performance issues. This can either avoid the need to send out technicians, thereby saving on truck rolls, or offers the ability to precisely pinpoint where the fault lies to prevent time being wasted trying to locate the issue.

How would AMI 2.0 boost our performance or our efficiency?

There is another piece in this puzzle: data analytics software. It wouldn’t be feasible to export csv files of all this data and hope to achieve something meaningful with it. However, there are companies out there, such as Future Grid, that provide software that can extract all of these data points and provide insights that could impact your operations.

Some examples include the following:

Fault detection: quick and accurate fault detection probably ranks right near the top of the list for most utilities. Not only does it make sure of the ability to meet performance requirements (and avoid penalties), but it also avoids the potential for costly downstream equipment failure and potential damage to humans or property. AMI 2.0 data can make fault detection around line down detection and ignition source detection more precise through more data points, which is a key aspect of wildfire prevention when it comes to grid equipment.

Load management: enabling dynamic voltage management and conservation, data analytics can also assist with Power Quality Management and load control. You can imagine the benefit to a utility of predicting and tracking consumer behaviors relating to heating or cooling depending on the time of year and climate.

Prediction for compliance and preventative maintenance: using AMI data to train the model, data analytics can help utilities understand how they are tracking against compliance (such as voltage tapped too high or too low) as well as performance issues signalling a requirement to investigate asset health.

Where does Ubiik fit into this AMI picture?

Ubiik offers an end-to-end AMI solution: from the meter read to the Head-End System (HES) and the downstream analytics solution. Providing a metering solution across a 3GPP standardized (non-proprietary) private LTE network opens the door to realizing some of the benefits outlined previously for AMI 2.0.

Founded in 2016, Ubiik started out as the world’s leading manufacturer of devices using Weightless™ LPWAN for industrial and national-scale, long-range IoT solutions. After being selected in 2017 as a vendor by Taipower (Taiwan’s national power supplier) the company has grown into the largest AMI solution vendor in Taiwan.

Ubiik’s release of the world’s first-ever LTE end-to-end AMI solution (meter NIC, base station, and NMS) supporting the ISM unlicensed band (915MHz) in North America has proved to be a game-changer for utilities. As a rare and increasingly expensive resource, spectrum often is the biggest barrier for utilities when it comes to private networks. Deploying a freeRAN™ pLTE base station allows utilities to gain the benefits of private LTE connectivity, even if they do not own or have access to their own dedicated spectrum.

Offering greater throughput than the typical RF mesh used for AMI, Ubiik’s pLTE AMI solution is ideally suited to support the high frequency polling and edge computing capability of AMI 2.0. With fewer base stations typically needed than a comparative RF mesh system would need (in gateways), the Ubiik solution offers a lower total cost of network ownership, with the added benefit of using the same communications network to support other utility applications.

Furthermore, Ubiik’s end-to-end meter communication solutions are differentiated by offering the ultimate flexibility in solving utility pain points. For example, we offer a single dual-mode meter Network Interface Card (NIC) to support both public cellular and a private network simultaneously. Regardless of whether a utility chooses to use public cellular or a private network as the primary connectivity, the same meter NIC always allows failover from one to the other, guaranteeing 100% coverage to meet stringent Service Level Agreements (SLAs).

How does Ubiik’s solution work?

The communications portion of the solution, the freeRAN™ base station, uses a robust private LTE Cat-M1 network, operating in the unlicensed 902-928 MHz ISM band. This private network is designed to cover the vast majority of meters managed by the utility.

Making sure of dedicated bandwidth and enhanced control over network management allows the Ubiik solution to offer improved reliability and security in meter communications. For areas where the private LTE network coverage may be insufficient or in cases of network congestion or maintenance, our system also seamlessly falls back to a Mobile Network Operator’s (MNO) public LTE Cat-M1 network. This dual-network approach makes sure of continuous connectivity and data transmission reliability across the service area.

Image 1: Ubiik Communication Architecture

My AMI mesh network has worked fine, why should I switch to pLTE?

Running AMI on a pLTE network as opposed to an RF mesh network brings a variety of benefits:

• Primarily, as noted previously, it’s about boosting the data throughput. This is a key requirement to support the bandwidth heavy requirements of AMI 2.0. The added benefit is that this allows any residual bandwidth to be used for other utility applications (such as SCADA).
• 3GPP standardization allowing for greater interoperability between devices (such as Distribution Automation equipment), as well as advanced security and interference mitigation technologies.
• The potential for a lower total cost of ownership due to typically fewer LTE base stations needed to support throughput than in a traditional RF mesh approach.
• Lower OPEX by reducing reliance on public cellular networks for connectivity.

Normally we’d keep all our data on-premises. Why would we consider using the AWS Cloud instead?

Transitioning to a cloud-based HES offers utilities access to real-time data. Instead of waiting for scheduled data transfers from legacy systems, smart meters now stream their readings directly through the HES into Kafka topics. This continuous flow of information opens the door to dynamic and sophisticated analytics. With meter reads available instantly, utilities can monitor consumption patterns, voltage fluctuations, and more, instantly. This immediacy is critical for operational scenarios that demand constant updates and rapid responses.

The cloud not only accelerates data collection but also simplifies the integration with a host of powerful analytics services. Traditional on-premises systems often struggle with data extraction and lengthy processing times. In contrast, the cloud-based HES seamlessly connects with robust suite of AWS analytics tools. For example, meter reads can be ingested into Amazon Redshift for scalable data warehousing or processed with Amazon EMR to handle large data sets. For those looking to harness the power of machine learning (ML), Amazon SageMaker offers an ideal platform for building and deploying models that can predict consumption trends or identify anomalies in real-time.

Furthermore, the AWS Meter Data Analytics (MDA) solution provides utilities with the tools necessary for forecasting and anomaly detection directly on the ingested meter data. This direct integration enables proactive maintenance strategies and improves customer engagement through data-driven insights. The overall result is a system that not only processes data faster but also provides deeper, actionable insights into customer behavior.

Are there cost benefits to this approach?

One of the key benefits of moving the HES to the cloud is the cost efficiency gained by processing data where it resides. Traditionally, moving large volumes of data from on-premises systems to a central processing unit is both time-consuming and expensive. With a cloud-based approach, utilities can bring algorithms directly to the data. This means that instead of incurring the costs and delays associated with moving massive datasets, data analytics and processing occur in the same environment where the data is stored. This streamlined approach reduces both latency and overhead, making it easier to scale operations and meet the demands of big data.

How did we deploy the HES on AWS?

Image 2: HES Architecture

In the diagram above, the (1) Ubiik LTE freeRAN base stations collect data from the smart meters and securely transmit the data through LTE to the telecommunication provider’s infrastructure. From there the data travels through a VPN connection to the AWS Cloud.

To make sure of high availability and optimal load distribution, incoming meter reads are handled by a (2) Network Load Balancer (NLB), which directs traffic to the (3) Ubiik deployed in an Auto Scaling Group. The HES processes these reads and performs two key operations: 1) it transfers the data to a (4) MySQL database running on Amazon RDS, and 2) simultaneously streams the data to (5) Amazon Managed Streaming for Apache Kafka (Amazon MSK) for real-time processing. The deployment process is automated and the application is containerized, thus the whole stack can be setup in minutes, which helps to onboard new customers more quickly.

There are several connections in that diagram. How secure is this process?

Security is paramount in this design. We’ve implemented VPC Endpoints for secure communication with both Amazon Elastic Container Registry (Amazon ECR) (for container image management) and AWS Secrets Manager (for credentials and sensitive data management). The MSK cluster operates exclusively in private subnets, eliminating the need for public endpoints and reducing the attack surface.

For downstream processing, we’ve implemented a separate VPC for the (6) MDA platform, connected through (7) AWS PrivateLink. This makes sure of secure and private communication between these critical components while maintaining network isolation.

For high availability, the HES with its database is deployed across multi-Availability Zones (AZs). This architecture not only provides robust security measures, but also makes sure of scalability and reliability for processing millions of meter reads efficiently.

Conclusion

The transition to AMI 2.0 represents a significant leap forward in utility management, demanding robust infrastructure to handle increased data loads. Ubiik’s pLTE solution, combined with AWS Cloud capabilities, offers utilities a comprehensive path to modernization. This partnership delivers enhanced security, real-time data processing, and scalable analytics while reducing operational costs. Using unlicensed spectrum and the powerful cloud infrastructure of AWS allows utilities to fully realize AMI 2.0’s potential, from improved fault detection to sophisticated demand response programs. The future of smart metering is here, powered by pLTE and cloud innovation.