Data Aquisition Word Cloud

DATA ACQUISITION SERIES PART II: CHOOSING YOUR SOFTWARE

DECEMBER 19, 2018

The first blog in this series addressed the importance of data acquisition, as it is the initial enabling step toward realizing the benefits of industrial data analytics. Choosing the right data acquisition software is critical to setting up your data acquisition system.

Data acquisition software makes it easier to track your data by writing the data files and linking them with important additional information, such as the date and time of the collection, the location of the sensor, the specific devices used, the data collection settings, and more.

When choosing a software system, you should ensure the application is compatible with your data collection hardware and your computer’s operating system. Many different developers offer data collection software, such as data acquisition hardware providers, like National Instruments (NI), or other third-party software solutions companies, like Predictronics. Software provided by data acquisition hardware manufacturers typically works seamlessly with the hardware provided, but it may not contain advanced functions, work with hardware from other manufacturers, or offer a friendly user experience.

Off-the-shelf Solution

One categorical choice would be an off-the-shelf solution. An example of this type of software would be PDX DAQ, an application within the Predictronics PDX software platform. Off-the-shelf data collection software is designed to be simple, intuitive, and easy to configure. These software solutions are designed for those who may not have specific programming knowledge and data science skills, but who still wish to monitor their devices.

Data acquisition software solutions provide methods with which the user can configure parameters for their data acquisition project. For example, data collection can be set to “trigger” at a particular time of day or at a point in which the data set reaches a certain value. Any combination of factors can be put forth as triggers by the user.

Triggering is a critical part of data collection, allowing the user to define the start and stop of the measurements. Without triggers the data acquisition software will endlessly and continuously save data, wasting vast amounts of computing resources and storage. This will give the user a large quantity of data, but it will not provide high-quality data that is specific to the user’s goals.

Other triggers can be set to check the measurement after it has been taken, utilizing statistical features to determine if the measurement is useful. This can cut down even more on unnecessary data processing, since useless measurements are discarded before being uploaded to a processing server. One example of this type of trigger would be setting a parameter to check that the maximum RPM of a CNC machine reaches a certain value when data is being collected. If the value is not reached, the measurement is discarded.

These applications are easy to use, but they may not be able to handle very specific tasks, such as computing specific features or creating exact tailor-made data collection routines.

Creating Your Own Solution

Another option would be to use a data collection software development platform. One notable example of this type of software would be National Instrument’s LabView application. Users of LabView can control almost every step of their project and customize it to their needs.

However, for some customers, this method requires more manpower and time than is economically feasible. This option also requires extensive domain knowledge and programming expertise in order to write your own data collection routines.

Depending on your application, it may be important to consider a software platform that has edge analytics capabilities. Edge analytics is an optional step after data collection where some features are extracted from the data. This lowers bandwidth usage, as less data needs to be transmitted over the network to a storage or processing server, thus helping to cut costs. The storage footprint of the features can easily be less than 0.05% of the footprint of the raw data.

Join us for the third and final part of our Data Acquisition Blog Series, Cleaning and Inspecting Your Data.


For more Predictronics information and updates, visit and follow us: