Pretesting Software Speeds Deployments

By Ken Traub

Before you go live with an RFID system, make sure your field and enterprise software are working the way they should.

You've developed a business case for using RFID to improve operations. You've chosen tags to track your assets, and fixed and handheld readers to capture the data. Now, while you're procuring the hardware, your IT department can develop the software components—and fully test them, before the readers are installed.

There are two software components: field software—middleware that captures data from RFID readers and outputs cleansed data—and enterprise software, which receives and integrates the cleansed data. If you design the interfaces up front and employ test harnesses, the two components can be tested at the same time.

First, design the data that flows between the field software and the enterprise software in each use case. This includes the cleansed RFID data to be used by your business applications. (To design this data, you can use the Electronic Product Code Information Services standard or a similar data model; see How to Deploy EPCIS.) It may also include queries the field software makes for enterprise data, such as shipping manifests or product masters.

Next, mock up the data for each use case. The mock data takes the place of the information you'd get from a live field system. Now, one IT team can use the mock data to develop any enterprise software that might be needed for new business processes, and to test how new and existing enterprise software receives and integrates RFID data.

Meanwhile, another IT team can develop and test the field software. For this, you will need an RFID reader simulator—software that mimics the behavior of an actual RFID reader. (Most middleware vendors provide reader simulation software.) A simulator lets you create virtual tags on a computer screen, which the simulator reads. With each tag read, the simulator sends data to your middleware using exactly the same protocol as a real reader.

Most reader simulators let you simulate missed reads and stray tags, and some let you record a complex sequence of tag reads and then play it back as you test and revise the software. You can use the simulator to explore many different real-world scenarios and confirm that the field-software outputs match the data for each use case you've mocked up. There are also simulators for handheld readers, and they include a simulation of the device's keyboard and screen.

The final step is to bring the field software and the enterprise software together to test the complete data flow, using the reader simulator in place of actual readers. Now, you're ready to go live with actual tags and readers, with the confidence that your software is up to the job.

Ken Traub is the founder of Ken Traub Consulting, a Mass.-based firm providing services to com­panies that rely on advanced software technology to run their businesses. Send your software questions to swsavvy@kentraub.com.