In this interview, Shailesh Mangal, the CTO of Zephyr, discusses the Internet of Things and how it changes the way we think about the devices we use. He also explains what sensory monitoring and smart devices are, and details how IoT devices have changed testing.
Josiah Renaudin: Today I’m joined by the CTO of Zephyr, Shailesh Mangal. We will be speaking on the Internet of Things and its impact on testing. Shailesh, thank you very much for joining us. First, can you tell us a bit about your experience in the industry?
Shailesh Mangal: Sure, Josiah. Thank you very much for having me. Zephyr is a test management company built to help our customers manage their testing lifecycle, be it cloud-based, on the web, or now the latest kid on the block, the IoT—Internet of Things testing. Those are tailored to meet our customers' needs—to take care of a variety of such use cases and help our customers test these in the field as well as in the labs.
Josiah Renaudin: Could you explain your personal definition of the Internet of Things? It’s a hot-button topic now. A lot of people are discussing it, and a lot of people are trying to explain what it is. What is Zephyr's definition of IoT?
Shailesh Mangal: Right. You said it well; I think there are so many different definitions floating around. The way we see IoT, essentially, is it's interconnected, uniquely identifiable computing devices that can be embedded and connected using the existing Internet's infrastructure. The important part here is the word “connected,” and it's not communication just from person to machine or machine to person, but also from machine-to-machine communication.
Josiah Renaudin: You had mentioned earlier ... we had talked a little bit before the call about how you work with a lot of different teams when it comes to testing devices. How has the Internet of Things impacted the world of testing? Have we had to rethink the testing process in this realm, or has the Internet of Things simply enhanced our previous testing processes?
Shailesh Mangal: Well, it's both. That community is leveraging from the existing experience of testing, so you still have the test cases, the test scenarios, the test plan that’s still applicable within this world of testing. We still need testers who need to continue and move this style of testing forward. What is new in this world is the introduction of test subjects. The test subject could be an active person, for example, to monitor or test a health device. It could be an animal to test a bio-chip. It could be a driver to test a particular automobile. You do need a specific subject in this world.
The second thing is the rethinking of the collection of this variety of data—testing will have to be both subjective as well as objective. Not everything can be measured in certain cases, as the testing has to go into the wild, so to speak, where we need to create the test environments to test the devices and put them through the real-world use cases. This could mean very adverse environmental conditions, varied degree of temperatures, very high- and very low-pressure scenarios, underwater testing—things of that nature. What changes is the way data is collected and the large amounts of data being analyzed. Systems in the back end analyze this large amount of data and inform the connected devices to take the right steps accordingly. To sum it all up, essentially there are some parts of testing that remain the same, and a bunch of these things can be leveraged from what we have today. A bunch of this data or process we will need to be rethink, though, and build out from scratch.
Josiah Renaudin: Absolutely. Can you touch just a bit on sensory monitoring and how the testers that you've worked with, the different partners or clients that you've had, how they've had to take sensory monitoring into account?
Shailesh Mangal: One of the important aspects of IoT, or one of the basic tenets of IoT, is the ability to sense something and record it. That could be the ability to smell, it could be pressure monitoring, temperature monitoring, depth, velocity … it could be any of these. This data essentially needs to be captured from the small devices and needs to be stored in the back end. There are a bunch of things that matter in this case. One thing is that these devices are usually very tiny and very low-capability devices. While testing these, one needs to make sure that the sensor data—which is communicated over to the main server if the connection is not available, or can also be stored locally on these devices and whenever the connection is available—can be offloaded back to the main server, so having offline capabilities is important. The sensor itself might be obstructed, especially in the wild when we talk about subjecting these devices to adverse conditions. The ability of the sensor to generate the accurate data may be compromised so the tester needs to take this into account and test it appropriately with proper boundary conditions.