In my recent travels throughout Europe, I witnessed numerous promising and innovative IoT technologies.
While we can all relate to wearables for sport and fitness, smart homes, and even “Siri” and her many step sisters, the real surprise is how the Industrial IoT (IIoT) is quietly gaining significant momentum among businesses and industries throughout the world.
Quite simply, executives are experiencing tangible results from their IIoT investments by way of decreased costs, greater profitability, and more efficient operations.
In other words, ROI for the IIoT is very real.
Growing pains persist
As with any transformative technology, there are growing pains along the way. The IIoT is no exception. There are connectivity challenges from cloud to gateway to end node. Not only must these devices connect to each other in a secure fashion, but they must also collect, aggregate, process, and send data up and down this secure IIoT infrastructure, which is composed of different wired and wireless physical media, communications and device management protocols, Cloud vendor solutions, etc.
With so much inherent complexity, it’s only fitting that we have some forms of fragmentation.
How do we solve fragmentation?
If we are to address fragmentation, and make IIoT adaption a more fluid and painless process, it’s important to discuss what’s required of next-generation IoT architectures. In this way, we can streamline current IIoT adaption – for businesses of all sizes – while we move forward with new and innovative solutions.
Through my interactions with industry executives and business partners, I’ve identified four general challenges for next-generation IIoT architectures. While the industry is addressing some of these requirements now, albeit in a narrow focus, those of us in the embedded systems industry must look at these requirements in a more holistic manner.
These four challenges are:
Comprehensive device management
Device management (Figure 1) is a core requirement for a successful IoT strategy. How do you manage all of your connected devices? Device management means having a strategy in place for complex operations such as device/OS updates, maintenance, application management, as well as fleet-based firmware rollouts. Of course, all of this must be done in a comprehensive and secure fashion north and south – across the entire IIoT infrastructure.
click for larger image
Figure 1: Device management must take into consideration all gateways, end node devices, and potentially multiple cloud backends. (Source: Mentor Graphics)
Multiple cloud solutions
There are a significant number of cloud vendors building compelling, but different solutions that enable IIoT. These vendors use standard communication protocols, which they enable with their embedded SDKs, but they use proprietary device management and application deployment and management solutions. IIoT equipment manufacturers must build devices that can support the multiple cloud environments in which their equipment will be integrated. This gives rise to a very important question: How do you cost effectively design and manufacture equipment that can be integrated with a variety of cloud environments?
Scalability and reuse
Because of the complexity and different types of devices, and the evolving topologies required to implement these devices, a cookie-cutter formula is not feasible. Invariably each device requires a different target hardware and software platform, which then requires a custom implementation. Moving forward it would be nice to have a solution that can scale across these disparate platforms maximizing code reuse and minimizing engineering resources.
Remote system health monitoring and diagnostics
A well-functioning IIoT system should be able to self-monitor itself for any potential issues, and then diagnose and repair them. When system degradation or failure occurs, system diagnostic functions are remotely executed to provide useful information to the IIoT operations team. Remote system health monitoring and anomaly detection improves system uptime while lowering costs related to maintenance, service, and equipment replacement.
The “productivity” gaps
Clearly, the four challenges I just mentioned are a lot to consider. We can summarize the challenges that they pose into two types of “productivity gaps.”
These two gaps are the IoT capability gap and the device implementation gap (Figure 2).
IoT Capability Gap: To varying degrees, cloud vendors invest to enable connected devices to take advantage of their cloud backend features. The investments range from providing a simple embedded library of connectivity protocols to enabling greater capabilities such as providing embedded SDK hooks for gathering telemetry data or initiating software updates. For most IIoT systems, the desired capability includes all of the above and more. However, the various cloud backend solutions only provide a fraction of the needs to satisfy the desired end device capability. The difference, or the gap, is called the IoT capability gap.
Device Implementation Gap: Not surprisingly, cloud providers do not focus their investments to broadly enable the specific runtime environments of all types of devices to connect to the cloud backend. How could they when connected devices range from simple, low-cost, single-purpose devices, to highly complex devices such as smart edge gateways capable of executing machine learning and artificial intelligence? These devices might run on a proprietary OS, an RTOS , or Linux . The required features of a given device determine how it will be implemented on a specific embedded platform. This includes integrating the aforementioned SDK hooks to the platform, implementing specific boot strategies for software updates, and instrumenting these devices to provide critical system health monitoring and diagnostic data, just to name a few instances. The device implementation gap can consume a significant amount of engineering and testing resources.
Figure 2: What’s needed to address the productivity gaps is a framework that complements and extends cloud vendor SDKs, and at the same time, enables portability of an IIoT system. (Source: Mentor Graphics)
Needless to say, business executives, software architects, and end-device manufacturers are all looking for powerful and more efficient ways to address these complexities.
There are ways to solve these challenges
Mentor has just announced a new framework approach that addresses IIoT fragmentation and the productivity gaps (Figure 3). The new framework called Mentor Embedded IoT Framework (MEIF) does not replace technologies and investments already provided by cloud vendors; rather, it fills the capability gap by complementing and extending those technologies, and fills the implementation gap by integrating the features fully with edge or end node device platforms. The new Mentor framework is both cloud and OS independent.
Figure 3: The Mentor Embedded IoT Framework complements and extends cloud vendor SDKs and enables integration and portability to the underlying device platform.
If we look closer, the MEIF design enables integration of cloud-vendor provided embedded SDKs (dark grey in Figure 3) alongside a well-defined set of IIoT-enabling runtime software (light blue in Figure 3), which can be extended as needed. Through this framework, users can connect all of their devices throughout the IIoT infrastructure in a secure, scalable manner. Costs are minimized when it comes to learning, implementing, and the porting of smart devices from powerful gateways to smart sensors on the edge.
Businesses that invest in IIoT are realizing the benefits. As a result, these companies are now implementing more complex and expansive next-generation IoT architectures (i.e., more edge and end node devices in more complex topologies). Device manufacturers and all parties involved are faced with challenges related to device management, unknown/multiple clouds, portability, scalability, and the need to remotely monitor and diagnose devices.
The Mentor Embedded portfolio, along with the Mentor IoT Framework, complement and extend the investments made by cloud backend vendors to provide comprehensive IIoT features and capabilities – right down to the hardware of an edge or end node device.
The benefits from using such a framework are clear: minimize learning curves, simplify implementations, increase code reuse, and reduce porting and testing costs.
For a more detailed explanation of the Mentor Embedded IoT Framework and our involvement in IIoT technologies, please visit the Mentor Industrie 4.0 website.
Scot Morrison is the general manager of the general embedded business unit, Mentor Embedded Systems Division, a Siemens business. Scot oversees the Linux®, Nucleus®, and Mentor Embedded Hypervisor runtime product lines, as well as associated tools, middleware, and professional services. Prior to joining Mentor in 2012, Morrison served as GM and SVP of products at Wind River Systems, Inc. Before that he worked at Integrated Systems Inc., where he last served as the VP and GM of the design automation solutions business unit in 1999, responsible for MATRIXx, and the pOSEK embedded operating system. Morrison earned his Bachelor of Applied Science degree in Engineering from the University of Toronto, as well as his MS degree in Aerospace Engineering at the Massachusetts Institute of Technology, specializing in control systems.