Device virtualization is key to IoT adoption

The Internet of Things (IoT) – a web of devices connected over the internet – included 9.7 billion devices in 2020 and is expected to exceed 29 billion by 2030. Bringing together the physical and digital worlds, the IoT is transforming every industry imaginable by opening up new opportunities; improving the customer experience; improving productivity, efficiency and agility; and enable insightful decisions.

Whether it’s deploying drones to survey farmland, using sensors and RFID tags to monitor goods through a supply chain, or delivering better banking experiences through connected user devices, the possibilities of the IoT are endless.

IoT must-haves

However, companies must meet certain requirements before they can fully use the IoT as a tool for business transformation.

First, the IoT must be embedded in products and processes, just like other software applications.

Second, success in IoT adoption is earned through iteration: Since the IoT has a myriad of elements, organizations must collect data from devices, turn it into analysis, act on that insight, and do it all over again in a continuous feedback loop that is always refining, learning and improving the various IoT components.

This implies that organizations must continuously manage, maintain and update the interconnected systems, processes and devices. And they have to test everything to make sure that all ‘things’ talk to each other and that their performance meets specifications and meets user expectations.

The following example shows a typical IoT testing scenario:

An instrument in a health care tracking system monitors a patient’s vital signs and records this information so that health care providers can access it when needed. Clinicians can initiate medication changes or intakes remotely from a computer or mobile device connected to the instrument.

To run smoothly, several aspects of this use case need to be tested.

For example, each device must be checked for usability (sends messages, logs data, displays information, etc.). All connected devices and the data that flows between them must be secure. It is essential to check the compatibility of the different operating systems, browsers, devices and connectivity options involved. The entire system must also perform at scale and meet all regulatory requirements. Furthermore, the software that powers the IoT devices must be thoroughly tested to eliminate bugs and optimize performance.

Since all these elements are scattered and under multiple owners, it is quite certain that they will not be active at the same time to undergo physical testing.

Returning to the earlier example, imagine the difficulty of physically testing a user interface that patients can access remotely from their respective (diverse) devices. Apart from the unavailability and inaccessibility of devices, the high cost is a major challenge on the device side when testing and validating an IoT solution.

Yet another difficulty in physically testing an IoT use case within corporate premises is that it requires massive resources that are only available in the cloud.

A strong case for device virtualization in testing

Under these circumstances, simulation testing based on device virtualization is a good option. Device virtualization — similar to creating a digital twin — addresses the aforementioned challenges by providing a layer of abstraction to IoT devices and systems.

The virtual machines simulate everything from device initialization, to two-way communication between devices and cloud, to manipulation of configuration settings. Different loads and network related scenarios can be simulated virtually to test the performance of an application. All types of devices, both in prototype and production phases, can be simulated through device virtualization.

In addition, using virtual devices (or the digital twin of physical devices) for testing reduces total cost of ownership and testing time.

Virtual appliances are particularly useful in the earlier stages of development, when their early feedback can be plowed back to eliminate bugs or fix performance issues earlier in the development cycle, and at a lower cost.

The benefits of device virtualization can be very significant. A financial services company shortened the feedback loop of its overnight regression cycle from 1500 hours on sequential testing to just 7.5 hours.

Last but not least, virtual appliances can automate 50 to 60% of testing requirements.

Forward-thinking organizations have moved beyond traditional testing methods and are making extensive use of virtual appliances and simulation in testing. Virtual simulation and feedback loops are an integral part of product development.

A good example of this is Dassault Aviation, which launched a business jet without making a physical prototype. Working on a virtual platform and shared database, the company’s global developer network has helped reduce assembly time and tooling costs by a very significant margin.

Improving the development results of IoT solutions

Device virtualization, in conjunction with IoT platform engineering, can also improve the quality and delivery of IoT solutions. The availability of highly capable, affordable devices is one of the driving forces behind the IoT revolution. So in addition to adopting new software innovations, IoT platforms must also keep pace with the evolution in hardware devices.

The problem is that hardware enters the IoT platform development cycle at a very late stage rather than at the beginning, resulting in higher costs, lower quality, and longer lead times. Device virtualization helps introduce hardware early in the platform’s development cycle – at the design stage of the application itself – and makes it accessible from anywhere. In addition, it benefits the development of IoT solutions in many ways.

For example, a virtual replica offers a way to solve a common problem in prototyping: parallel development of hardware and applications, which may make a physical device unavailable during integration testing. The virtual device replica steps in, mimics the new features, and provides feedback to speed up device prototyping.

Two other scenarios where virtual devices add value are platform engineering – where they help with programming and compliance testing – and ensuring that IoT applications are compatible with previous, current and future versions of various devices.

Last but not least, device virtualization improves feature validation and test results by testing a range of parameters including scalability, resource utilization, and security.

Balakrishna DR, popularly known as Bali, is the executive vice president and head of the AI ​​and automation unit at Infosys.