Example: Testing the efficiency of new motion sensors by using an analytic rule

Emily, the energy manager at Supermarket XYZ, uses IBM® Maximo® Real Estate and Facilities Real Estate Environmental Sustainability Impact Manager to monitor and improve energy efficiency. To test the efficacy of new motion sensors on an aisle of refrigerator units, Emily creates an analytic rule.

Background

To reduce energy use, Supermarket XYZ installed motion sensors to control the lights inside an aisle of refrigeration units. The refrigeration unit lights are off by default to conserve energy and are automatically turned on when the motion sensor detects movement.

Emily wants to use an analytic rule to test that the new motion sensors are actually leading to lower energy use during periods of low traffic in the aisle. Her logic is that the power usage of the unit should not be high when the light device energy use is low. In other words, if there is low traffic in the refrigerator unit, the doors should remain closed, and it should require less power to keep the refrigerator cool.

Emily plans to create the analytic rule and then monitor the events that are triggered for a two week evaluation period before determining if the motion sensors are effective enough to invest in for more aisles.

Step One: Create a plan for an analytic rule that captures the correct information

Emily's first step is to answer some questions that will help her to determine what is needed to created an analytic rule that will test the motion sensors' efficacy.

Table 1. Emily's plan
Question Answer
What is the goal? To test that energy use is low when there is low traffic. Emily wants an analytic rule to alert her each time the refrigeration units' power usage is higher than she expects during low traffic. To determine what periods qualify as low traffic, Emily uses the amount of power that is used by the lights. Based on historical data, she determines that energy use below 35 watts indicates few people in the aisle.

What is the device type or classification?

Does the device type already exist at Classifications > Specification Class > Operations (Specification Class)?

The device type is Refrigeration Units.

No Refrigeration Units classification exists at Classifications > Specification Class > Operations (Specification Class).

What are the data points that are needed?

Do the data points already exist in the hourly fact table?

 The data points needed are:
  • Power usage in watts
  • Motion light sensor in watts
The Asset Analytic Hourly fact table (triAssetAnalyticHFact) includes a data point for Power Usage In Watts (triFactPowerUsageWNU). But there is no data point for motion light sensor usage.
What is the asset analytic rule logic?

Based on these answers, Emily determines that the analytic rule logic should be:

Refrigeration Unit is running, Power Usage In Watts > 1500W AND Motion Sensor Light In Watts < 35W, Met for 2 hours

Step Two: Create the analytic rule and associated workflows

To create the analytic rule as well as the needed device type and data point, she must complete the following tasks:
  1. Create the device type "Refrigeration Units" by creating a specification class.
  2. Create a data point for the motion sensor light, which is measured in watts, in the Asset Analytic Hourly BO (triAssetAnalyticHFact).
    1. Based on the naming guideline, Emily names the data point cstFactMotionSensorLightWNU. The cst indicates customer field, Fact indicates that it is a fact data point, MotionSensorLight is the field, W indicates the unit of measure is watts, and NU indicates that the field type is number.
    2. For the label, Emily enters Motion Sensor Light in Watts. Emily does not select the Do Not Auto Populate check box.
  3. Create a new analytic rule. Emily creates a rule named cstRU00001 with a priority set to High.
    1. For the description, Emily enters Refrigeration Unit is running, Power Usage In Watts > 1500W AND Motion Sensor Light In Watts < 35W, Met for 2 hours.
    2. She decides on the name of the rule engine workflow that she will create for the rule and enters it into the rule as triAnalyticRule - Synchronous - RU - cstRU00001 - High Power Usage in Low Traffic.
    3. For the event summary workflow, Emily enters triAssetEventSummary - SubFlow - cstRU00001 - RU - High Power Usage in Low Traffic.
    4. She chooses to have the rule trigger a notification rather than an asset event request. As a possible resolution, she enters Refrigeration Unit is not sealed, door is not closed, or unit required maintenance.
    5. Finally, she adds the parameters: “Power Usage In Watts” with value “1500” “Motion Sensor Light In Watts” with value “35”.
  4. Create the rule engine workflow triAnalyticRule - Synchronous - RU - cstRU00001 - High Power Usage in Low Traffic.
    1. In the first branch, for the Retrieve Properties Task, Emily enters into the Label field Get Power Usage in Watts Param.
    2. In the Filter Using section, Emily clicks General::triNameTX and sets the value for the Right Data field to Power Usage In Watts.
    3. The, for the Define Variable Task, Emily enters Power Usage in Watts Param into the Label field.
    4. Emily confirms that the For Task field is already updated with the renamed task Get Power Usage In Watts Param.
    5. Emily sets up the second branch to include the Motion Sensor Light In Watts parameter.
    6. Emily checks that both the Power Usage In Watts parameter and the Motion Sensor Light In Watts parameter have a Results Count value of 1.
    7. Emily sets the conditions to be Power Usage In Watts > Power Usage In Watts Param && Motion Sensor Light In Watts < Motion Sensor Light In Watts Param.
  5. Create the event summary workflow triAssetEventSummary - SubFlow - cstRU00001 - RU - High Power Usage in Low Traffic.
    1. The workflow that Emily copies has four branches to capture four rules data points, but Emily only has three data points defined. Therefore, Emily removes the extra branch and its tasks.
    2. In the first branch, in the Query Task, on the Filter Using section, Emily clicks General::triNameTX and sets the value to Power Usage In Watts.
    3. For the description, Emily enters Create Asset Event Summary Intermediate Record for Power Usage In Watts.
    4. On the Edit Map, Emily enters Power Usage In Watts as the value for triNameTX and General::triFactPowerUsageWNU as the value for triParameterValueNU.
    5. Emily verifies that the value is Power Usage In Watts (triAssetEventSummaryIntermediate).
    6. Emily sets up Motion Sensor Light In Watts in the second branch.
    7. Emily sets the value on the third branch to Is Running.

Step Three: Create associated reports

Emily creates a system report query to filter the hourly fact table for the new device.

  1. Emily names the report Cst - triAssetAnalyticHFact - RUs and enters Asset Analytic Hourly - Refrigeration Units for the Header (Title).
  2. In the System Filter Columns section, Emily modifies the Specification Class (triDimSpecificationClassTX) row. Emily enters Specification Class in the Report Label field, sets the Filter Operator to Equals, sets Conditional to No, and enters Refrigeration Units into the Value field.

Step Four: Create and run the analytic process

Emily creates an asset analytic process with the new filter and the new analytic rule and periods. She uses the ID Cst - RU - 2 Hours Rules Process” (RU for Refrigeration Units) and the name Cst - Refrigeration Units - 2 Hours Rules Process. For the description, Emily enters Analytic Process to execute Refrigeration Units rules that evaluates 2 hours periods of Asset Analytic Hourly reading. Emily chooses Cst - triAssetAnalyticHFact – RUs as the system report query in the Criteria Query Name field. Emily adds the rule she created, cstRU00001, as the only rule to be included in the process. Finally, Emily schedules the analytic process to run on a regular basis for the two week evaluation period.

Step Five: Monitor the results and evaluate the strategy

If events are created when the process runs, Emily receives a notification. She monitors the events and looks for possible causes. At the end of the two weeks, she finds that with the sensors installed, energy use is lower when there is low traffic in the aisles. She uses the data she has collected to make a case for installing additional sensors.