Data Aggregation at the Inport

SpeeDBee Synapse:v4.9.9Last updated:2026-09-17

Purpose

Combine data received from multiple components with slightly different nanosecond timestamps into one record for each configured window width, then write the records.

Project files

These are the project files used in this case study. A standard version and a version with the PLC Collectors configured in simulation mode are available.

• InportAggregation.sspj
• InportAggregation_Sim.sspj (simulation mode)

Note: Import requires SpeeDBee Synapse version 4.9.9 or later.

What you will learn

• Why data from multiple components is output as separate records
• How to use the aggregation function in the inport settings to combine data into one record

Requirements

Collect temperature and pressure from each PLC with PLC Collectors connected to PLC-A and PLC-B, and input the data into the same inport of the CSV Serializer. Divide the input data by the configured window width and combine the data within each window into one record.

Component	Data	Address	Description
PLC Collector (PLC-A)	temperature	D0100	Temperature (°C)
PLC Collector (PLC-A)	pressure	D0101	Pressure (kPa)
PLC Collector (PLC-B)	temperature	D0100	Temperature (°C)
PLC Collector (PLC-B)	pressure	D0101	Pressure (kPa)

Even when PLC-A and PLC-B collect the same types of data, their collection timings do not completely match. Because timestamps are assigned with nanosecond precision, the slight timing difference causes the timestamps to differ and the records to be separated.

When aggregation is not configured, the CSV Serializer outputs data as follows. Because the PLC-A and PLC-B data have different timestamps, each record contains data from only one PLC.

timestamp	PLC-A:pressure	PLC-A:temperature	PLC-B:pressure	PLC-B:temperature
13:08:04.947378445+0900	68	25
13:08:04.949666433+0900			82	31
13:08:05.947378445+0900	68	25
13:08:05.949666433+0900			83	31

Note: The date portion (YYYY-MM-DD) is omitted.

Because data with different timestamps is output as separate records, use the aggregation function to combine data within the same window into one record.

Components used

• PLC Collector (collects data from PLC-A and PLC-B)
• CSV Serializer (converts aggregated data to CSV)

Workflow

1. Collect data from PLC-A and PLC-B with the PLC Collectors
2. Input the data collected from PLC-A and PLC-B into the CSV Serializer
3. Use the aggregation function in the CSV Serializer’s inport settings to combine data within each configured window into one record
4. Output the aggregated data as one CSV record

Setup

1. Configure the PLC Collectors. Specify the connection target, addresses to collect, data types, and other settings for PLC-A and PLC-B so that temperature and pressure can be collected. In this configuration, collect temperature from D0100 and pressure from D0101 on each PLC.

   

   

   Strict timestamp

   When Strict timestamp is OFF, data within the same component is assigned the same timestamp. When it is ON, even data collected by the same PLC Collector is assigned different timestamps, so the records are separated. In this case, it is turned OFF because data within the same component should have a common timestamp.

2. Configure the CSV Serializer. In this case, set the file switch count to 10 so that one file is output for every 10 records.

   

3. After configuring the CSV Serializer, connect each PLC Collector to the CSV Serializer’s inport. Configure the flow so that data from multiple components is input to the same inport.

   

4. Right-click the CSV Serializer’s inport, open the inport settings, and turn Aggregation ON. Specify the Window width. In this case, set it to 1 second and combine data within each window interval (1 second) into the same record.

   

5. Configure aggregation for the input data. In this case, specify the target using regular expressions. Set the component name to .* and the data name to .* to target all data input to the CSV Serializer.

   

6. To output only aggregated values, turn off RAW DATA. Click the aggregation toggle for the specified input data and select newer from the aggregation options. In this case, only the newer value within the window is used.

   

   

   Aggregation options

   newer is used in this case. However, when collecting data at short intervals such as 100ms, or when setting a larger window width so that many data items are included in a window, mean and other options can also be selected as representative values. Select an appropriate aggregation option according to the purpose of the data.

7. Save the settings and start each component.

   

Checking the output

Data within the same window interval is output as one CSV record.

Download the aggregated CSV sample

The following is an excerpt from the CSV output after aggregation. Even when timestamps assigned to data from each component differ at nanosecond precision, data within the same window can be output as one record.

timestamp	PLC-A:pressure.newer	PLC-A:temperature.newer	PLC-B:pressure.newer	PLC-B:temperature.newer
14:37:13.551084040+0900	71	26	86	34
14:37:14.551084040+0900	71	26	87	34
14:37:15.551084040+0900	71	27	87	35
14:37:16.551084040+0900	72	27	88	35
14:37:17.551084040+0900	68	25	82	31
14:37:18.551084040+0900	68	25	83	31

Note: The date portion (YYYY-MM-DD) is omitted.

Sub-millisecond timestamp values

The sub-millisecond portion of the timestamp is determined based on when the Synapse component starts processing. Therefore, the sub-millisecond portion cannot be aligned to zero in the output.