Appliance - Presence Discovery
The Appliance - Presence Discovery subject area includes data elements about whether a major appliance is present at a customer's household. This data is automatically generated by Oracle Utilities proprietary data science models. The models use customer AMI data and weather data to predict the presence of an appliance at a household.
The goal of this information is to help utility program managers improve the performance of their energy efficiency programs and marketing use cases. For example, you might use this information to send targeted communications to your customers to promote an electric vehicle program, or to offer a rebate incentivizing customers to install a more energy efficient type of electric heating.
On this page:
Usage Notes
Treating the Presence Discovery output as a prediction
The output of the Presence Discovery models is a prediction, not an absolute. This means that some households might not be identified as having a major appliance, while other households may be misidentified as having a major appliance when in fact there isn't one. Use caution depending upon the intended use.
For example, if you intend to use this data to send targeted customer communications, think about how to adjust the language of your communication to account for the possibility of a misidentified home, and offer a clear way to opt out of the communication if it isn't relevant to the customer. You could also design your communication to prompt customers to take the Home Energy Analysis survey so that they can confirm the presence or absence of a major appliance, and thereby receive more accurate personalized recommendations.
If you have any questions or would like some guidance about how best to use or interpret this information, contact your Delivery Team.
Using the Presence Discovery subject area versus attribute
The Appliance - Presence Discovery subject area is based on timeline data and includes data for all past historical weeks, as defined by the International Standards Organization (ISO). It should therefore only be used for historical analysis, such as showing how the number of major appliances has changed over time. For all other scenarios, use the shared Presence Discovery attributes, which are tied to customer IDs and can be used in other subject areas. For example, you can use the shared Presence Discovery attributes in the Household - Count of Customers subject area to see how many major appliances are predicted to be present in a specific city. See Pre-Built Dashboards - Disaggregation Insights for more examples of such use cases.
Data Elements
Data Element | Description |
---|---|
Appliance |
The type of appliance that may be present at a customer's site. The appliances currently available for presence discovery include:
Note: Solar customers are excluded from certain presence discovery processes. This is because the data science models have not yet been trained to produce results for solar customers who have Level 1 EV chargers or electric heating. Therefore, customers who have Level 1 EV chargers will be excluded if the presence of solar technology is predicted to be "very likely" for them. Similarly, customers with electric heating will be excluded if the presence of solar technology is predicted to be "very likely" or "somewhat likely" for them. Type: Attribute |
Appliance Detections |
The number of total detections of all or selected appliances. Type: Attribute |
Site Presence |
The likelihood that an appliance is present at a customer site. Oracle Utilities uses proprietary data science models to determine the likelihood of an appliance's presence. The values currently available are:
Type: Attribute |
Response Count |
A count of the total number of responses to the Home Energy Analysis. Type: Measure |
Unique Household Count |
A count of how many unique households have one or more major appliances. For example, this shows a count of how many households at a utility are very likely or somewhat likely to have electric heating, or how many households do not have electric heating. The households in this case are households with AMI data. Type: Measure |
Limitations
- Non-residential customers are excluded from the Presence Discovery process.
- Customers with more than one electricity service point are excluded from the Presence Discovery process.