The future of IoT testing

  February 28, 2017

The Internet of Things could turn out to be the biggest software development and testing opportunity of all time. While IoT devices such as phones, tablets and wearables may already seem ubiquitous, the additional vast and networked array of embedded IoT systems and sensors could eventually outnumber these familiar portables with a worldwide reach of expanded applications and services.

The Internet of Things brings unprecedented scope and new challenges

A 2015 Business Insider Intelligence report estimated that by 2019, the Internet of Things could receive a larger public reception than the smartphone, tablet, PC, wearables, and connected car markets combined. Enterprises are expected to be the most eager adopters of IoT initiatives, outpacing both governments and consumers. En route to integrating more than 23 billion devices into verticals such as manufacturing and retail within the next four years, IoT development is expected to thereby significantly increase the scope of Internet of Things testing.

The Internet of Things poses fresh challenges to developers and software testers that will stretch their skills in releasing top-notch, quality-assured software. In a way, the coming years will be a replay of what happened after mobile devices took off in the early 2000s. Then, test engineers had to account for factors such as battery level, available bandwidth and current network coverage that had been less pressing with traditional desktop and laptop computing.

IoT likewise brings many new things to the table in terms of testing requirements, including a major emphasis on unusual device types, such as home appliances, minimal thermostat interfaces, and watch sensors. With a diverse set of products entering the IoT market, QA teams must carefully chalk the best test management strategy to effectively meet the quality standards of these products. Let's look at what the future of IoT testing may hold for software testing teams:

The importance of testing wireless connectivity scenarios

Connectivity throughout the IoT is dependent on many different wireless standards. For example, to do anything at all, an IoT-enabled medical device might have to interface with Wi-Fi, Bluetooth, ZigBee and/or 4G LTE. On top of that, hardware may have to deal with electromagnetic interference while also complying with strict medical regulations.

For software makers, the potential issues with connectivity and infrastructure will inevitably shape the design of their applications. Software testing will need to cover bases such as what happens to data when a connection is unexpectedly dropped. Is it saved and properly stored? Where will it pick up once service is restored? Plenty of real-world testing - e.g., walking around in an RF-saturated environment while switching between connections - will be essential.

Testing IoT devices in a controlled lab environment is different from testing the same devices under conditions into which they will be deployed. The diverse interfaces involved in real time usage of IoT devices require connection to dissimilar networks and placing reliance on dependable two-way signal management. Whether devices are communicating with a server or between each other, uniquely designed assurance of continuous data transfer is fundamental to IoT operations.

Many IoT devices are out of network detection range. But devices that are connected to a network need to be detected as being on or off line. Online detection is also followed by monitoring the device condition and level of operation.

Empower the device with the capability for extended power consumption. Efficiency in power usage requirements is a worthwhile investment in the design and development IoT devices.

Service virtualization for simulation of smart homes

One of the most touted IoT use cases is home automation. Expansive bandwidth is necessary for the expanded reach of IoT devices for such functions as home automations, automatic door locks, and their consolidated operations. Devices such as Dropcam and Nest have already made waves with consumers, while frameworks such as the Apple HomeKit are rolling out an attempt to weave together the different parts of the connected home.

For testers, however, home automation can be a tough and unfamiliar environment to simulate. Considering compatibility, “What other devices are present?”, or coverage, “What's the layout of the home?” can be daunting exertion. Fortunately, service virtualization offers one possible way forward. Dev/test teams can model many different types of houses, sensors and device states. Testers can get a good sense of what conditions their services will face in the real world.

Doubling down on security in the IoT

Although the IoT appears to have an enormous upside for today's businesses, its potential downsides are worth noting. In particular, security lapses almost always come up in conversations about what could go wrong with a system as vast and heterogeneous as the IoT. A 2014 study from HP estimated that 70 percent of IoT devices were vulnerable to attack. The difficulty of patching everything in the IoT only compounds this issue - the components of the IoT are not as straightforward or uniform as PCs and smartphones.

Securing the data and data privacy is crucial IoT connectivity. Authorizations are important supports for securing IoT data streaming and transfers. For instance, to avoid data vulnerability, the IoT device port is closed to Internet communication when not in use. In addition, end-to-end encryption between devices, or devices and servers, securely codes data transfers.

QA teams must vet a full range of potential vulnerabilities in IoT products and services. Depending on the item in question, this diligence could take the form of enforcing strict password rules, protecting the interface from unauthorized access and ensuring the use of encryption where appropriate. The HP report found that transport encryption and insecure Web interfaces were among the top issues with the IoT.

Testing for a wide range of interfaces

With devices like Dropcam and Nest, software testing is relatively straightforward since the hardware is well-known beforehand. Other services such as cross-platform cloud-based software present more complicated scenarios for quality assurance teams, since the client device could be almost anything.

With video streaming sites such as Netflix, accessible testing already exists on anything from an HDTV to a handheld gaming console. In these cases, it may be useful to construct a minimal set of requirements that can be ported and tweaked depending on the device. Usage analytics could also provide insight into which devices merit the most time and energy. Along the way, knowledge of the relevant APIs and updates will be crucial for performing worthwhile testing.

There are several appropriate test options for IoT devices, from the granular testing of chips and sensors to test coverage for multi-platform data flows. Calibrated stimuli can cover chips for measurement and observation of such properties as the electrical signal shape and strength, memory length, and display accuracy and appearance.

An IoT, such as an automated door locking system, is supported and coordinated by a central server. The testing of IoTs connected to computing systems involves multiple units and requires broad coverage, depending on the type of device being tested and the type of testing being performed. In addition to electrical tests, devices can require tests on acceleration, vibration, stability, repetition, stress, memory, or endurance. Creative solutions are required for testing the Internet of Things, often causing longer test times.

What’s next?

Across the board, the Internet of Things could restructure software development and release into an intuitively in depth operation. The amazing growth in customer demand for IoT is amazing software enterprises, with many predicting that by 2020 the Internet of Things will permeate the industry. While businesses have been the largest initial consumers of IoT, public enthusiasm in mobile computing devices has recently outdone enterprise interest. The current challenge is consequently to increase the scope of IoT testing.

Automated testing is crucial to designing, planning, and implementing IoT testing. With a whole new set of development and testing requirements IoT is set to be a reminder of when mobile phones and tablets took off in the early 2000s. IoT brings new development and testing needs to software deployment.

Testing IoT devices requires metrics able to read the sensors and impulses that regulate home appliances, smartwatches, thermostats, exercise trackers, pacemakers, and car door locks. The phenomenal diversity of the independent computing devices about to hit the software stage is anticipated to further engage customers with enumerable applications available at a touch.

Along with real time testing, simulation models are tremendous resources for IoT testing. The reasonably extensive number of IoT connectivity levels requires that testing tools and procedures verify numerous communication protocols, including WiFi, Bluetooth, CDMA, and 3G. As well, data transport requires continuous security that involves access controls, authentication, encryption, and private access. Additionally, different IoT devices have dissimilar construction, with each IoT design bound to have unique testing requirements. Simplicity in IoT design and construction facilitates testing of IoT devices.

The real time environments in which IoT devices must operate exposes them to weather, elements, and physical impact. Consequently, QA teams must test the full range of both IoT environmental and functional vulnerabilities, including to what extent exposure affects functionalities. Among the primary IoT testing challenges is transport encryption.

Conquering the business landscape to stake an enterprise claim is the strategic goal of successful organizations. Mitigating risk in product deployment conquers the worst case scenario of a failed, incomplete, or otherwise poorly deployed release. QA testing of IoT devices that is innovatively, comprehensively, and flexibly constructed to align with product design will go a long way towards mitigating risk, reducing time to market, and ensuring ROI in alignment with enterprise needs and priorities.

The Internet of Things is an assortment of dedicated physical devices containing embedded technology that interacts with the internal state of the user and/or the external environment. It is therefore useful to devise a set of test requirements and analytics essential to the efficient performance of the specific device to be tested. IoT is structured with communication, applications, and data analysis for monitoring and managing a variety of ‘things’. The variety must be built into IoT testing protocols.

Looking at the innovative onslaught of the Internet of Things one sees the advancement of technology into mobility and diversity. With test integration and automation, QA teams are well-set to devise testing procedures which will secure the further expansion of IoT technology.

Related Articles: