Testing the IoT

By 2020, more than 50 billion devices will be digitally connected, representing $19 trillion in business opportunity. As systems increase in size and complexity, engineers and scientists face new challenges when networking and synchronising devices to make intelligent decisions.

Although the Internet of Things may sound like marketing jargon to some test managers, smart sensing devices are poised to become the centre of a new technology revolution. Industry players such as Microsoft, Raytheon and Qualcomm have already achieved distributed sensing, intelligence and analytics through differentiated software and hardware platforms. For progressive test managers, IoT is the opportunity to achieve rapid improvement, higher throughput and lower test costs.

The Internet of Things (IoT) is reaching the test and measurement industry at an interesting time. Sceptical test managers may have heard about the next tech disruption, but some may have a hard time recognising how the technology could impact their systems. The same test managers have been at the helm of distributed sensing and analytics technology for decades — the same technology IoT proponents proclaim is driving progressive change.

The technology may seem familiar to automated test organisations and there is significant opportunity for forward-looking test managers to profit from this megatrend. The key is to understand how the IoT applies to test and measurement, and learn from established companies that are using a platform-based approach to benefit from this technology. By using a proven software and hardware platform, test managers can reap dividends of productivity today while preparing their organisations to thrive tomorrow.

So, what is the IoT? Usually it refers to systems that were formerly treated as isolated entities but are now sources of relevant data. Some examples in the consumer space include: smart appliances, thermostats and power meters.

The IoT is also changing industrial processes and business operations — this is the area where test managers can make a difference. The tools used in the Industrial IoT (IIoT) range from test and measurement instruments such as oscilloscopes, multimeters and function generators to fully integrated smart testers. The secret ingredient that takes these tools from mere isolated things to IIoT-enabled devices is a combination of technologies: device-to-device communication, automated analytics and scalable systems. Fortunately for the test and measurement specialists, this is not uncharted territory. Industry leaders have implemented these technologies through software platforms for more than a decade.

Creating a device-to-device communication network

At release in 2006, NI LabVIEW 8.20 software was notable for including a web server for building test and measurement systems. Why would someone automating data acquisition need a web server? For automated test developers at Microsoft working on testing Xbox 360 controllers, this technology presented an opportunity for unprecedented device-to-device communication. Combined with GPIB, serial and other network communication protocols such as TCP/IP, a web server provided a way for test systems to communicate results to each other and back to a central archive.

When developing new functional tests for the Xbox 360 controllers, developers found that optimising as many parallel tests as possible in a limited production cycle time was a significant challenge. Microsoft realised that knowing the status of all of its testers and viewing a central repository of all test data presented an operational advantage to optimise these tests and produce less expensive, more reliable devices. Microsoft is not alone in this pursuit; test organisations around the world are building software infrastructure to remotely monitor, analyse and even control their production test systems.

Deriving value from test data through automated analytics

Similar to how infants babble to each other, for a long time test devices have muttered unintelligible command line statuses through serial ports. Device-to-device communication is truly valuable when meaningful insights can be derived from the data through automated analytics. This is an area where the test industry has long been ahead of other industries.

For aerospace giant Raytheon Missile Systems, wide varieties of analog data are logged to a central repository of binary files tagged with relevant metadata to allow the data to be analysed. Automated analysis of this data is conducted with tools such as LabVIEW and NI DIAdem to gain a high-level view of the performance and immediately report results back to operators and managers. Without a software platform to manage this data and generate useful insights, test organisations will be overwhelmed by the substantial amounts of collected data and they’d be unable to derive value from it. In Raytheon’s case, however, the insights gave the company the ability to reduce the time required for each test cycle by half.

In addition to centralised analysis on the server, for distributed test systems, automated analysis on the node can provide significant operational advantages. Although raw analog data may overwhelm a network, an FPGA or CPU processing on the node can synthesise data into fragments such as average values that can be more succinctly communicated over the network to other stakeholders. Equipped with a high-level graphical programming approach called the LabVIEW reconfigurable I/O (RIO) architecture, test engineers can take advantage of user-programmable FPGAs and embedded controllers to perform distributed analytics or make instant decisions at the node without needing to send data back to a central server.

Planning for changing requirements

In addition to device-to-device communication and automated analysis implemented in software, another crucial aspect of IIoT systems already found in today’s test and measurement applications is scalable hardware systems. For wireless telecom leader Qualcomm, disparate, lengthy measurements on traditional box instruments were driving up the cost of measurement in an extremely competitive and cost-sensitive industry. By upgrading from RF signal generators and analysers to NI’s modular, PXI-based hardware platform, Qualcomm saved significant rack space and lowered the cost of test. More importantly, by taking advantage of the user-programmable FPGA hardware on NI’s vector signal transceiver, Qualcomm decreased test time by a factor of 200.

Building on a flexible hardware platform that uses the latest commercial technology such as NI PXI turns future requirements from a threat of obsolescence to an opportunity to upgrade. Instead of replacing an entire box instrument for a new measurement specification, a new PXI module can be incorporated into the existing test system. Or if the processing power for a complex RF measurement is inadequate, the PXI controller can be replaced with the new NI PXIe-8880, which features an eight-core Intel Xeon processor. This flexibility opens up test systems to exciting new technology enablers, such as silicon advances, user-programmable FPGAs, and timing and synchronisation advances.

IIoT systems are constantly evolving to include more and more nodes with varying I/O requirements. To meet these challenges, a software platform must support flexible, modular hardware solutions.

Business opportunity in the IIoT

Successful companies have used a proven software and hardware platform to implement elements of the IIoT.  Microsoft built a networked software infrastructure to connect its testers and gained insights to double test throughput per station. Raytheon implemented automated analysis of data to reduce test cycle times by 95%. Qualcomm radically redefined its definition of test coverage through a modular, software-designed instrument approach.

Although the IIoT has already manifested itself in test and measurement over the last decade through software platforms such as LabVIEW, today’s test managers are presented with significant new opportunities to make their test systems smarter. Web, cloud and mobile technology are redefining how humans interact with machines and machines with each other. For the first time, fully automated end-to-end testing, analysis and adaptation are real, tangible possibilities.

For some test managers and organisations, there may not be a choice. As test budgets are further squeezed, requirements continue to change rapidly and time-to-market pressures increase, piecemeal hardware systems with software built from scratch will become increasingly untenable.

LabVIEW and NI TestStand deliver the software platform needed for test engineers to use scalable test hardware such as NI’s PXI hardware platform to achieve device-to-device communication and automated analysis. With a proven approach for defining acquisition, synchronisation, processing and analysis through high-level graphical programming, NI developed LabVIEW 2015 to help engineers write code faster. With NI’s software portfolio, test engineers can spend more time focusing on the test challenges at hand without getting caught up in the low-level details.

This empowers developers to build a robust IIoT solution while avoiding the risk of building a software solution from the ground up. LabVIEW facilitates device-to-device communication through countless methods such as PXI chassis backplane communication buses, embedded network protocol support and HTTP web methods. LabVIEW and TestStand facilitate automated analysis wherever it needs to happen — on the instrument, on the tester or on the data server. Innovative technologies like NI Cloud Dashboard further extend test devices to be able to seamlessly upload streams of test data to the cloud where they can be accessible from any device with an internet connection.

This software platform puts control into the hands of test engineers, while empowering them to use hardware with industry-leading commercial technology like NI’s new eight-core PXI controller or high-voltage system SMU.

Returning to the original question is the IIoT a disruptive new opportunity or just validation of proven technologies in the test and measurement industry? It is up to each test manager to decide. What is certain is that the IIoT will give rise to smarter test systems that will further redefine test and measurement in the decades ahead.

Image credit: ©iconimage/Dollar Photo Club