SoMs speed the move to the Industrial Internet of Things

by ERIC MYERS, National Instruments

Systems-on-modules and platform approaches can help keep IIoT efforts from getting lost in the weeds of interfacing and hardware development.

Engineers who design equipment for the Industrial Internet of Things (IIoT) will likely face a number of technical challenges. One of the most significant is in developing systems that are adaptable and that can scale with the IIoT.

To get a feel for the problem, consider the aircraft maker Airbus and its experiences developing a Factory of the Future. This is a long-term research and technology project that employs emerging computing and communications technologies to build aircraft faster, more flexibly and with higher quality. Airbus plans to develop many systems—such as smart tools, inspection devices and robotics—that will connect and work in harmony to improve the overall manufacturing process.

Like most traditional design firms, Airbus began by designing its concepts from the ground up. It eventually recognized that connecting all of these systems in a smart way is not trivial due to the many communication mechanisms and protocols involved in an IIoT network.

As Airbus learned, the task of building a complete system from the ground up takes a substantial amount of time. During the initial design, teams spend most of their time making components work together. Only a small amount of time is spent developing the special functions of individual nodes.

Another challenge lies in scaling these systems to grow with the expanding IIoT. When such a system is first deployed, it generally works well. But nodes in the network must be able to change and adapt and new devices are constantly added. It’s not unusual for new devices or functions to force a redesign of the network from the ground up. For example, were Airbus to add a new robotic system on its factory floor, the move would force a redesign of other systems to support the proprietary protocol involved.

It’s also not uncommon to see teams developing systems using multiple off-the-shelf subsystems, sometimes developing parts themselves, then using off-the-shelf devices where appropriate. In machine control, for example, many systems today add health monitoring capabilities using off-the-shelf subsystems.

This approach speeds development, but off-the-shelf subsystems are typically closed and fixed. Closed architectures tend to have a limited ability to expand. If they can expand at all, it is generally though one or two expansion ports that use a proprietary design, perhaps requiring a license fee from the manufacturer. And it may take technicians with specialized tools or training to install any enhancements. The proprietary nature of closed designs tends to limit the amount of information they can share over the network. Moreover, data that can’t be communicated through an open standard interface can’t be analyzed by other devices, eliminating one of the benefits of the IIoT.

The way around this difficulty in the IIoT is by deploying a network of “things” flexible enough to evolve and adapt. Teams need an evolved approach that focuses on the innovation within the application itself, not on hardware or software. This is known as a platform-based approach, one emphasizing systematic reuse of software and compatible hardware, intended to reduce development risks, costs and time to market.

One example of a platform technology that is becoming widely used is system on module (SoM) and computer on module (CoM). For hardware developers, an SoM provides the processing, memory, peripherals and I/O elements needed in any IIoT system. For software developers, an SoM comes with anything from a board support package (BSP) to a complete software suite fully supporting the hardware and connectivity to other systems.

In the case of Airbus and its Factory of the Future, the firm decided to adopt NI SoM, providing core hardware components along with a software suite to support the board. Airbus estimates that switching to this approach cut its development effort by a factor of ten.

Among the entities developed for the Airbus Factory of the Future are smart tools. A given airplane subassembly has about 400,000 fastening points that must be tightened. Human assemblers handle the task using over 1,100 different tightening tools. The operator must follow a list of steps and verify the proper torque law settings for each location. To eliminate possibilities for error, a smart tightening tool uses machine vision to understand the tightening task at hand and automatically set the torque. The outcome of the task gets recorded in a central database. This lets production managers review procedures and processes during quality control and certification.

With a platform-based approach and growing technology though, these teams will be able to efficiently develop the IIoT, such as Airbus with the NI SoM.

References

Airbus Factory of the Future case history
sine.ni.com/cs/app/doc/p/id/cs-16246

LabVIEW overview
www.ni.com/labview

Systems on Modules
sine.ni.com/nips/cds/view/p/lang/en/nid/212787