2016 IoT & Embedded Technology The Real Need for Real-time Java Gateway Solutions Exclusive License to Distribute: PTC By Chris Rommel, Executive Vice President
1 Introduction The Internet of Things (IoT) is forever transforming the embedded landscape. Connectivity is rapidly redefining system functionality and challenging existing concepts of differentiation. Cars are now assessed as much on their in-vehicle infotainment stacks as they are on their transmissions. Meanwhile, the field of industrial automation is focusing on leveraging connectivity to predict maintenance needs and promote operational efficiencies. While the largest impact of the IoT might be its basis as a catalyst for change with existing applications, it also is fostering the creation of entirely new device classes, such as the gateway, that evolve from the need for new points of data aggregation and intelligence within the IoT network topology. In many instances, these evolving use cases and emerging device classes require fundamental shifts in the underlying device technology to support IoT product functionality or business goals. For example, one of the most apparent layers of change has been with the operating system. Powered by semiconductors with more on-chip memory, more systems are adopting formal operating systems when prior designs used in-house operating systems or had no OS at all. In fact, 75% of respondents to VDC s 2015 IoT and Embedded System Engineering Survey reported using a third-party operating system on the current project, compared to 46% of respondents in 2008. Implicit with the transformative nature of the IoT, however, OEMs new business goals often require more than just a third-party OS they necessitate more robust connectivity middleware stacks. In some cases, engineering organizations are obtaining this functionality through the new operating systems themselves, but in others it requires additional investment, either in supplemental software or in the form of additional labor investment to create the middleware in house. This pressure to develop new software extends beyond middleware and is especially acute in the application and services layer. With software content creation demands accelerating, engineering organizations are looking for new ways to augment their productivity. In many cases, this is translating to investments in new development tools and programming languages. Further, the demographic base of the available labor pool is shifting. Today, professional skill sets and academic curricula have evolved to center on higher-level object-oriented languages such as C++ and Java. Accordingly, engineers skilled in embedded development stalwarts such as C have become increasingly scarce and expensive. Background on VDC Research VDC has been covering the embedded systems market since 1994. The findings from VDC s 2015 IoT and Embedded System Engineering Survey, capturing the input from over 800 engineers, offer insight into leading business and technical trends impacting engineering organizations as well as the best practices implemented to address them. The respondents are directly involved in software and systems development
2 across a range of industries including automotive, aerospace and defense, telecoms, medical, industrial automation, and consumer electronics, among others. Gateways Pacing IoT Market Growth Exhibit 1: Forecasted Global Shipments of IoT & Intelligent Gateway Hardware by Revenue (Thousands of Dollars) Gateways are one of the fastest growing and developing device categories in the embedded market with annual growth rates approaching 15% per year. What began as hardware intended to serve as a localized point of presence for data aggregation has rapidly evolved to address the growing computational and intelligence requirements of the Internet of Things. The related gateway use cases have likewise expanded from rudimentary remote monitoring to more advanced data analytics and context-aware real-time control (see Exhibit 2). While supervisory control and data acquisition-like (SCADA) use cases still exist, increasingly intelligent edge devices are redefining and expanding needs for gateways in IoT network topologies. Many of these new use cases require distributed, remote control and bilateral communication. Therefore, they often necessitate more sophisticated gateway hardware and software solutions that can support intelligent device management, such as over-the-air updates, device security, services provisioning,
3 and so on. Additionally, gateways can help standardize the fragmented and evolving wired/wireless technology and lower-level protocol preferences of industrial systems as well as lessen the exposure to changing requirements and the risks associated with them. Exhibit 2: Most Attractive Application Services Deployed through IoT Gateways (Percent of Respondents) *Note: Sums to greater than 100% due to multiple responses.
4 The rapidly growing amount of data being created by IoT devices reinforces the need for more intelligent gateways within many operational technology deployments. For one, more gateways are being used as a mechanism to provide localized data storage and/or analytics. While many initial gateway deployments focused on the simple aggregation and transmission of data from multiple sources, many enterprises are now searching for ways to intelligently analyze and/or filter data prior to communication to the datacenter. In many ways, the gateway can become a localized extension of enterprise or cloud-based analytics engines, optimizing edge system performance for environmental conditions based on ongoing neural-net learning. The emergence of real-time embedded processing and control capabilities on intelligent gateways thus could lessen the importance of communication bandwidth and datacenter storage requirements. As more organizations pursue the IoT opportunity, the ability to offer post-deployment services is central to the vision of IoT providing a mechanism to develop new revenue and customer engagement models. Gateways are also accelerating the adoption of cloud services for industrial IoT applications. Sixteen percent of respondents indicated they are currently deploying or using cloud services through gateways today, and another 26.6% of OEMs and other end users expect to do so in the next 12 months. We expect this trend to continue as more organizations recognize gateways as critical components of IoT service enablement strategies that can offload computational burden from edge devices and provide necessary localized analytics. More Ecosystem Evolution Required The need for more sophisticated gateway solutions comes at a time when the hardware supplier ecosystem is still maturing. In fact, more than two-thirds of respondents indicate that their organizations are doing at least some internal development for or creating IoT gateway systems. Many of the leading IoT and intelligent gateway vendors develop these sophisticated systems after historically manufacturing smaller modules and similar communications devices such as routers and modems. This lineage has often led to the production of gateway solutions that lack the level of integrated software and application platforms that many deploying organizations would prefer to help minimize the amount of internal integration and development labor that has become the norm.
5 Exhibit 3: Source Used or Expected to Use for IoT Gateway Technology (Percent of Respondents) Today only 27% of respondent organizations are completely outsourcing integrated gateway systems. Moving forward, we expect the majority of gateway solution providers to continue to build their hardware systems internally from the ground up. The use of third-party solutions will increase going forward, however, fueled by the growing availability of reference hardware for software/service providers to leverage and build solutions. While more organizations may look for turn-key gateways as the market matures, the growing value of software-intensive use cases will drive the need for better solutions targeted at that domain. For example, more developers will be looking for real-time solutions for their gateway devices. To date, many of the initial vendor solutions and support offerings are centered on Standard Edition Java running in a general-purpose Linux environment. While Linux is important and a logical point of prioritization, the demand for real-time data analytics and decision making calls for some additional alternatives. The increasingly diverse use cases for gateways necessitate a range of run-time options and, in similar fashion to the operating system landscape, a heterogeneous assortment of choices with some geared at real-time and/or deterministic functionality.
6 IoT Increasing Java Value Proposition Embedded software development was once predominantly conducted in C. However, embedded engineering has evolved. Embedded systems are no longer synonymous with small-footprint, fixed-function devices. They are more sophisticated, more intelligent, and more connected. This evolution has likewise fueled a migration to new programming languages. Just as developers once migrated from Assembly to C, engineering organizations are already embracing alternatives. In fact, the use of C has dropped to almost half of its previous adoption levels with more engineers now opting for object-oriented and scripting languages like Java and JavaScript. Exhibit 4: Languages Used to Develop Software on the Current Project (Percent of Respondents) *Note: Not all language choices are depicted above. Sums to greater than 100% due to multiple responses. The benefits offered by object-oriented and higher-level languages such as C++ and Java have outweighed those offered by C, which had maintained a strong hold on embedded development. The magnitude of this change over just five years is tremendous given the traditional pace of evolution in the embedded market, further highlighting the overall acceleration in the rate of innovation and new feature set requirements. In
7 essence, today s software content creation demands are placing further pressure on the productivity of embedded engineers. The reuse of existing assets is a critical part of organizational development best practices. Java not only provides extensive middleware libraries, but Java 8 s release in 2014 included Lambda expressions, which can provide additional efficiencies by letting engineers avoid some boilerplate code and further reduce development time. Additionally, the widespread use of Java for enterprise/it development can help ease the integration of new IoT and cloud services with existing enterprise infrastructure and applications. This broader use of Java outside of the traditional embedded market offers additional value to OEMs. Engineering organizations can leverage the broader resource pool of IT/ISV developers. Not only is the use of the C language decreasing as previously discussed, but the traditional embedded engineering population is simply too small to serve IoT market development needs. The simplification of both hardware and software development languages is allowing OEMs to tap into the broader population, which is nearly 10 times the size of the traditional embedded engineering community. As mentioned earlier in this document, the laws of market economics yield significant related labor cost savings. Exhibit 5: Project Schedule Adherence, Segmented by Software Development Language Use (Percent of Respondents)
8 Our research showed that the median fully-loaded cost of an engineer working on a project where Java is used is 10% less than the overall industry median cost ($95,000 versus $105,000). Furthermore, the average Java projects are shorter in duration and require fewer engineers than those for C and C++, which could translate to additional time-to-market and cost savings. These Java projects are also more likely to be on schedule, with 83% of engineers using Java reporting that their project schedule adherence improved versus their previous project (see Exhibit 5). As more organizational revenue is tied to post-deployment services, project length more directly impacts revenue and can even risk missed market windows. Thus, the ability to choose platforms that enable flexibility in resource selection will place access to more complementary third-party software at a premium. Change in Status Quo Required The growing utility and application of these new software development languages, however, cannot completely displace the need for solutions catering to the needs of systems with real-time latency requirements. In fact, 60% of engineers state that their current project requires some real-time capabilities. Even newer device classes such as gateways require real-time functionality for some of the more automation/control -oriented use cases. Gateway performance also can be of elevated importance when considering the growing volume of data generated from edge devices that necessitate some form of rapid and localized analytics for context-aware behavior. As discussed, many traditional barriers and reservations about using Java are wearing down too. Latency fears, driven by Java garbage collection, originally caused many embedded engineers to discount the potential of Java for real-time applications. These concerns have largely abated, however. Already, a majority of surveyed embedded Java users are using Java in a project with some real-time requirements. In fact, the percentage of Java developers reporting real-time capability requirements is almost exactly the same frequency as the embedded engineering population as a whole. There will most certainly always be a sub-set of real-time projects that require lower hard real-time latency performance, safety-critical certification, or whose system resource constraints simply preclude Java s use. While Java may not be the best fit for those applications, we believe that the proportion of relevant real-time projects addressable by Java will continue to increase going forward. An ecosystem of solutions targeting real-time use cases for Java actually already exists and has been servicing this space for over 15 years. The growing use of Java has not yet universally been paired with a commensurate understanding or awareness of best practices for its implementations, however. Although commercial solutions such as PTC s are also available, many engineers would be well served to explore the potential need for such a solution, especially in light of the growing use of intelligent gateway devices for the IoT.
9 Summary & Conclusion The IoT has unquestionably accelerated the evolution of traditional embedded device classes. In some cases, the related changes are incremental. In others, however, they can be revolutionary and even category redefining. One such device type is gateways, which is not only a quickly growing device class, but also an increasingly common part of IoT network topology architectures used to enable intelligent systems of systems. Organizations are recognizing their utility for data filtering, analysis, and distributed monitoring and control. Meanwhile, the aforementioned market changes are also causing many OEMs to reevaluate their incumbent development technologies. Java, for example, continues making inroads for IoT and embedded development given the combination of changing developer skill sets, device resources, and functional requirements. The growing experience developing with Java in embedded engineering organizations is only reinforcing this growth trajectory even within applications for which it was once deemed ill-suited, such as those with real-time requirements. The convergent growth trajectories of Java and gateways within the IoT market will make the technologies increasingly common pairings in the future. Next-generation implementation differentiation, however, will rely on solutions that can offer a degree of specialization that can better address the range of heterogeneous IoT use cases. And as use cases and IoT service offerings evolve, the value of real-time analytics and system situational awareness will be placed at a premium. This dynamic will require the adoption of similarly specialized sets of solutions, such as those targeted at real-time applications, for their gateway devices. Engineering organizations that can recognize those places where specialized and premium solutions can add value will be at a decided advantage within the hyper-competitive IoT marketplace.
10 VDC Research About the Author Chris Rommel is responsible for syndicated research and consulting engagements focused on development and deployment solutions for intelligent systems. He has helped a wide variety of Contact Chris: clients respond to and capitalize on the leading trends impacting crommel@vdcresearch.com next-generation device markets, such as security, the Internet of Things, and M2M connectivity as well as the growing need for system-level lifecycle management solutions. Chris has also led a range of proprietary consulting projects, including competitive analyses, strategic marketing initiative support, ecosystem development strategies, and vertical market opportunity assessments. Chris holds a B.A. in Business Economics and a B.A. in Public and Private Sector Organization from Brown University. About VDC Research Founded in 1971, VDC Research provides in-depth insights to technology vendors, end users, and investors across the globe. As a market research and consulting firm, VDC s coverage of AutoID, enterprise mobility, industrial automation, and IoT and embedded technologies is among the most advanced in the industry, helping our clients make critical decisions with confidence. Offering syndicated reports and custom consultation, our methodologies consistently provide accurate forecasts and unmatched thought leadership for deeply technical markets. Located in Natick, Massachusetts, VDC prides itself on its close personal relationships with clients, delivering an attention to detail and a unique perspective that is second to none. For more information, contact us at info@vdcresearch.com.