By now you have probably heard this phrase everywhere and you may be a bit confused. What is smart meter analytics? How could it be useful to my utility? Where would we even begin?

This page aims to answer these questions and connect you to other resources to get you started. AMI meter data contains vast amount of actionable insights - if it can only be treated in a way that recognizes the unique features of utility/AMI data.

1. What are smart meter data analytics systems anyway?

2. What kind of data do AMI meters generate?

3. How much data do AMI systems create

4. What could my utility do with all this data?

5. What systems exist to handle the vast amount of data involved?

1. What are smart meter data analytics systems anyway?

Broadly speaking, smart meter analytics systems aim to generate actionable information from the vast amount of energy consumption and operational data that AMI smart meters provide. Many of these systems also function as a "system of record" which drive billing, maintenance and other key business functions. A true smart meter analytics system should have the following core capabilities:​

2. What kind of data do AMI meters generate?

AMI smart meters primarily capture time series energy consumption data in kilowatt-hours (kWh) on a 1-minute, 5-minute, 15 minute or 60 minute schedule. The most most common for residential meters is 15-minute or one hour. Many larger accounts will have data recorded at 1 or 5 minute intervals. Some brands of meter store the data as cumulative kWh reads, while other store the data as calculated kWh energy consumption. Cumulative reads need to be processed in order to yield energy consumption amounts.

AMI meters also store demand data which tracks the maximum kW demand in a given time frame (usually monthly). Different utilities will choose to track this data in different ways, depending on billing needs etc.

Voltage data can help grid operators regulate the grid to ensure safe and reliable power. The voltage data is not used for any billing or customer facing function.

Meters also create and send specific error codes back to the utility. These error codes will vary between the different manufacturers, however the common themes are the tracking of power outages, reporting on mechanical errors within the meter, attempted tampering errors and communication errors.

Lastly, AMI meters store meta information about the themselves. This meta-data can include identification codes (e.g. serial number), locations (both lat/long and address), customer information (account identifier and customer name, for example) and change histories (e.g. installation date).

3. How much data do AMI systems create?

Number of Meters	1-hour intervals	15-minute intervals
100,000	876 million	4 billion​
500,000	4 billion	16 billion
1,000,000	9 billion	36 billion
2,000,000	18 billion	72 billion
5,000,000	44 billion	175 billion

AMI (Advanced Meter Infrastructure) meters can generate a vast amount of data.​ The exact amount of data will vary from utility to utility, and from system to system. AMI meters are capable of recording electric or gas consumption values at almost any interval (down to seconds). Most electric applications use 15-minute or 1-hour intervals, with 5-minute intervals common for larger, commercial customers. Gas meters rarely collect data at more than a 1-day interval. Here is a rough guide to how many data points you can expect to collect on an annual basis:

Of course, consumption values are only one of a suite of values that AMI meters can collect. Information on voltage and error codes can quickly multiply the amount of data that needs to be stored for analysis. 

4. What could my utility do with all this data? 

Okay I get it - AMI meters create a LOT of data. Besides billing, what can we actually do with it? Here's a list just to get you started:

• Evaluate the performance of energy efficiency and demand response programs - or EM&V (Evaluation, Measurement and Verification)

This list isn't even exhaustive. The potential scope for the useful application of AMI analytics reaches every part of the utility industry and our ability to improve the way we do business will only grow as new applications appear. There are even organizations springing up specifically to help utilities handle this space.

5. What systems exist to handle the vast amount of data involved?

So now what? What can my utility do to get started? How do we make a plan for 

Tier 1: Traditional databases
​
Tier 2: Meter Data Management (MDM) Systems
 
Tier 3: Generic big-data solutions

​Tier 4: Utility oriented big-data​​
​