Driving the IoT Journey: 10 Trends to Watch INTRODUCTION The Internet of Things (IoT) market continues to gain momentum both growing in total connections and into new business and consumer application segments. Overall connection growth in 2015 was 26% for things connected over short-range wireless technologies and using cellular, satellite, and fixed line access services. The average net new IoT connected devices per day grew from 640K EOY 2014 to 1 MM EOY 2015. While these are impressive numbers, more impressive is that the IoT market is still relatively immature: in a 3Q 2015 survey of IoT end-users and suppliers asking them the development stage of enterprise IoT projects, the results showed that 70% of companies were either in the investigative stage or early design phase. But the IoT journey for all companies will be shaped by the future. In fact, some of the innovations set to change the IoT landscape were introduced over the past year. Demonstrations of augmented reality (AR) technologies showed how valuable IoT data becomes when it is visualized using a smartphone or even smart glasses. New IoT data and analytics offerings came to market, such as platforms that speed application development and automate predictive and prescriptive analytics. IoT supplier segments also changed to better serve customers by specializing, adding new services, or offering innovative new products such as IoT marketplaces. This whitepaper intends to show how these technologies and trends will impact the IoT market for the rest of 2016 and into 2017. It will also project some of the more interesting new opportunities and challenges that will shape the future of IoT.
ANALYTICS AUTOMATION SEPARATES THE LEADERS FROM THE FOLLOWERS Changing Model Algorithm New Data Analytics on collected IoT data is the goal of most, if not all, connected things and products. In analysis by ABI Research, by 2020, businesses will spend more than 26% of their entire IoT solution cost on technologies and services that store, integrate, visualize, and analyze IoT data, nearly twice what is spent today. However, analytics is still a very manual activity, particularly in the more valuable predictive and prescriptive analytics services. At the onset of any IoT analytics project, data scientists have to spend time building, testing, and refining data models. But because machines age and business processes evolve, as well as availability of new types of contextual data, data model adjustments become a periodic activity. Based on ABI Research data, the very manual nature of data scientists activities are seen in analytics professional costs that are expected to remain over 33% of enterprise IoT data and analytics spend through 2021. What can improve data scientist efficiencies and lower the enterprise professional services costs is automation of many of the more manual activities i.e., analytics automation. However, this is not an easy proposition because it requires a platform that can easily ingest new data, make adjustments to analytics models in real time, and even automate prescriptive analytics. Analytics automation is in its infancy today, but its impact will be profound. The efficiencies it provides to data scientists effectively democratizes analytics, allowing more problems to be solved and analytics use by more employees and business units. Analytics automation separates the leaders from the followers; it lowers the total cost of analytics ownership and raises the bar on bestin-class offerings. IOT EXPERIENCES - AUGMENTED REALITY DRIVES THE NEXT PHASE OF ENTERPRISE GROWTH IN IoT AR is the superimposition of data in the form of text, graphs, images, and even video, onto a live image of the physical world. Facilitated mainly through smartphones and tablets, most AR applications today serve consumer markets. But the real revolution in its use and value will be in enterprise markets when AR technologies are paired with IoT and other application data. Technicians are the near-term beneficiaries of AR and IoT. A technician can view a machine through a smartphone or tablet and quickly find parts, view use history, gain instructions on removal and installation, and even order the parts. AR and IoT applications also facilitate 2
efficient knowledge transfer to new workers. And when paired with wearables, such as smart glasses, additional synergies are gained by freeing workers hands. Smart glass applications will be a major driver for overall enterprise AR adoption, expanding value to new occupations, including surgeons and nurses in healthcare, and firefighters and ambulance workers in public safety markets. When combined with IoT data, and wearables technologies, AR s potential lies in its ability to personalize our interactions with the physical world effectively creating IoT experiences. Dashboard controls in cars, trucks, and other equipment can be augmented or even substituted with digital overlays. Workers in manufacturing environments can build personal workspaces for tool use, job tasks, and parts location. In smart home markets, AR technologies can assist seniors in finding necessities and medications. Today, AR supplements our visual senses but future AR technologies could serve all human senses including tactile, auditory, and olfactory (smell). ABI Research believes that this market is set to take off; enterprise AR software and services revenues are forecast to exceed $12 billion by 2020. INDUSTRIAL IoT MAKES STRIDES IN STANDARDS BUT THE SMART FACTORY IS STILL AN OEM BATTLEGROUND IoT is providing industrial markets, such as oil and gas, utilities, transportation, and manufacturing with the technologies and services for their next phase of growth. IoT is a core component driving the vision of the next industrial revolution, called Industry 4.0. In Industry 4.0, industrial processes and the associated machines become smarter and more modular, driving higher levels of utilization and greater flexibility for meeting customer demand. Achieving the vision of Industry 4.0 requires developments of standards, and steady progress has been made. OPC-UA, standardized by the IEC in 2011 and recently updated in 2015, is a vendor-independent protocol that allows diverse pieces of control equipment to communicate with each other, effectively enabling a hyper-connected network across multiple industrial ecosystems. The AvNu Alliance working group is driving development of the Time Sensitive Networking standard, a low-latency industrial communications protocol that can interface and deliver data over standard IEEE 802 IT networks. Other initiatives that will contribute to smart industrial systems include OneM2M, which provides a common service layer for monitoring and updating device and sensor software. The Bluetooth Special Interest Group is developing Bluetooth mesh that will enable robust wireless communications for industrial sensor networks. However, the benefits offered by these promising standards are challenged by key issues that characterize many industrial markets. Equipment replacement and upgrade cycles are long; industrial firms are slow to swap out equipment or augment networks if legacy equipment runs sufficiently and can meet production KPIs. Equipment in industrial settings has not typically been designed for sensor augmentation or even additional forms of connectivity to enable standard IoT services such as condition-based monitoring. As well, due to security and operational reasons, many pieces of industrial equipment lack ports for extracting processing data for analysis. 3
These challenges become even more pronounced for manufacturers seeking to evolve to the smart factory. The reason is manufacturing involves multiple processes from raw material ingestion, to material processing, assembly, and packaging, each served by multiple OEMs using their own proprietary solution. As a result, ABI Research expects that over the next five years, IoT connections in manufacturing will grow, but primarily from OEM solutions siloed around their equipment. In this environment, successful IoT suppliers targeting industrial markets will be those that help OEMs build their own IoT initiatives as specialists, but at the same time also offer services and platforms to bridge the various siloed IoT solutions. One-stop shop IoT suppliers are well-positioned to capitalize on both sides of the opportunity; however, only if they can satisfy two conditions: their services can be disaggregated and offered as micro-services; their specialist services are best-in-class. In this way, OEMs and manufacturers can more easily participate with the goal that early successes will lead to customers integrating into a broader supplier portfolio. SPECIALIST ONE-STOP SHOP APPLICATION ENABLEMENT PLATFORM ECOSYSTEM EVOLVES INTO SPECIALISTS AND ONE-STOP SHOPS Application enablement platforms were created to ease extraction of data from a device or machine, transmit the data efficiently over a network, and get it into a form that is easily consumable by an IoT application. In a world of legacy devices, proprietary protocols, and a highly fragmented thing market of sensors, devices, and machines of all levels of sophistication, AEPs bundle a suite of capabilities to simplify this integration effort. AEPs effectively try to solve the IoT plumbing issue, blending the worlds of OT and IT. Today, AEPs fall into two groups: specialists and one-stop shops. Specialists are those that are focused on one or a few services, such as messaging or application management typically agnostic to vertical market. Specialists also include AEPs that offer a complete set of services, including application development; however they tend to specialize in serving particular industry segments. One-stop shop providers are integrating either up or down the stack seeking to become the lead in an IoT engagement. There are two sub-groups in this segment. The first subgroup is suppliers who are offering hardware, connectivity, and AEP services. These suppliers eliminate the task of selecting a communication module or gateway supplier, as well as the telco for connecting the assets. Companies selecting these suppliers then use their own resources or outsource services for application development, analytics, and systems integration. 4
The second group offers AEP services plus other value-added services, such as application development, data management services, analytics, and other consulting and professional services. More advanced companies in this group will bundle their services into a broader IoT partner portfolio, including access to hardware and connectivity via an IoT marketplace. An important benefit of suppliers in this group is that products and services are pre-integrated, which eliminates many of the integration activity time and costs, facilitating faster time to market. The significance of this evolution is threefold. It demonstrates a healthy marketplace as suppliers try to innovate via technology development or offer expansion. Second, it shows that the thing ecosystem continues to expand, needing suppliers who are expert in not only extracting data from any thing but also merging data from different thing ecosystems. Finally, it demonstrates the complexity enterprises feel when combining the physical and digital worlds. AEPs, both specialists and one-stop-shops, are on the front line of this march to a smarter connected world. THING IDENTITY AND MANAGEMENT SERVICES As the base of things in IoT grows from tens of billions to hundreds of billions, it is inevitable that things will want to participate in multiple ecosystems, interact with other things, and be accessible to new services. We already see IoT device and services suppliers in smart home, connected car, and mobile healthcare beginning to tap into each others markets. Critical to facilitating these opportunities is thing IDENTIFICATION CONNECTIVITY SERVICE MANAGEMENT identity. Thing identity facilitates creation of thing management services and other new value-added services in such areas as security and analytics. Developing thing identity services requires network addresses provided by the up to 7.9 x 10 28 IP addresses of the IPv6 protocol. The second critical component is a thing s description and what it can do the thing s attributes. Companies and suppliers building IoT solutions today are already defining the attributes of their devices, machines, and things to both manage them and create services. But two challenges arise from these siloed efforts. The first is that the descriptions of attributes for things will develop non-linearly and vary significantly. How one manufacturer describes the function of its switch will be different from another manufacturer. The second challenge is that the types of things getting connected will also grow, compounding the variation in attribute types and descriptions. There are already initiatives to help address these issues. The physical web, where anything can have a web address, is one option. IoT frameworks from the likes of the OIC and AllSeen offer options for describing things to facilitate interactions across diverse thing ecosystems. But these tools may not simplify interactions between legacy and new devices, nor be acceptable to enterprises that closely guard access to their connected products. 5
Under these conditions, there is a strong case to be made for development of third-party services to help manage a variety of connected things for authentication, access, software updates, Over-The-Air (OTA) management, and security, as well as expose and translate device attributes. All these opportunities will be facilitated with thing identity and management technologies and services. THE RISE OF IoT MARKETPLACES One of the biggest challenges facing the IoT market today is supplier diversity and offer complexity. There are literally hundreds of suppliers to choose from to create an IoT solution. The numbers of SIs and VARs that help integrate the pieces of an IoT solution are vast mainly because they tend to be companies of 50 employees or less. Connecting the assets can be provided by a range of technologies, which are offered by hundreds of hardware vendors. The middle part of the value chain is equally complex with more than one hundred suppliers offering AEP services. Not only is the IoT supplier base vast, but even finding a singular entity to help navigate the technology and supplier choices is a challenge, as discovered in a recent ABI Research survey. Survey respondents were asked what are the challenges for IoT end-users with IoT professional services? Thirty-seven percent stated lack of suppliers with end-to-end technology know-how; 30% stated lack of suppliers with knowledge of vendors and partners serving all parts of the IoT value chain. Some of the more forward-looking IoT suppliers are addressing this challenge through a new offering called IoT Marketplaces. IoT Marketplaces aim to provide enterprises with a portfolio of hardware, software, and services to simplify assembling an IoT solution. A key benefit of the more advanced marketplaces is components and services that are nearly plug-and-play. Pre-integrated components are a major benefit because they tighten overall solution security and enable fast time-to-market. IoT marketplaces are relatively new, with only a few offering a full set of products and services to build, run, and manage an IoT solution. However, ABI Research expects this channel to grow as suppliers seek to take complexity out of the possible. 6
IoT DATA GOVERNANCE AND EXCHANGE SERVICES IoT data sharing is a relatively uncommon activity across enterprise markets. Most activities around enterprise IoT data are for internal activities of governance who, what, why for access and disposal; and for storing, processing, analyzing, and visualizing IoT data. While enterprises are far from exposing and sharing their IoT data with external parties, it is worth exploring the opportunities and challenges. The best example of IoT DATA EXCHANGE THIRD PARTY thing data sharing and its impact on all markets is thing locations exposed in mapping software. Digital access to the locations of roads, cities, towns, places, and businesses has facilitated the development of applications that we use every day. The most recent example is the development of the billion-dollar online transportation access industry, which includes companies like Uber, Lyft, Arro, and others. Interestingly, taxi location in these apps is not a shared piece of data outside of the applications themselves. However, vehicle location data sharing will need to become more accessible if the world hopes to achieve the potential for self-driving cars and intelligent transportation systems. Enterprise markets, particularly industrial, will be slow to expose their IoT data for reasons of security, IP protection, and market competitiveness concerns. But as their IoT business models mature, enterprise markets will look to understand how they can expose their data to improve their own operations and generate new revenues. Clearly, most enterprise IoT data would need to be anonymized, but doing so and then exposing it to a huge developer community could lead to an avalanche of new services, including marketing programs, support services, and new analytics services. Like thing identity and management services, the potential for data governance and exchange services will be handled best by third parties. Their services may be as simple as providing outsourced IoT data governance services. Or they can be as extensive as data exchange services that anonymize data ownership, translate meta-data, and act as settlement agents handling both the financial and operational aspects of data sharing. IoT data governance and exchange services are a longer-term opportunity, but the potential is huge for taking IoT to new levels of value and market collaboration. 7
EDGE COMPUTING EDGE COMPUTING EDGE COMPUTING CONTINUES ITS MARCH TO RELEVANCE The norm today with data and analytics services is that most IoT data is sent back to a cloud environment from which it is either stored or acted upon using business rules or sophisticated analytics. For some industries, regulations require storing activity and use data of their machines. For others, deep analysis of IoT data is used to create new products and services, as well as perform predictive and prescriptive analytics, both of which require cloud computing power and access to other data. However, market forces both on the supplier side and from customers are driving more use of edge computing in which edge analytics are applied on the device itself DATA CENTER or on the network edge. Edge computing helps limit the amount of data transmitted back to the cloud, which reduces connectivity costs over metered networks, and data storage/integration costs. In some applications, edge computing is needed because the application cannot wait for an analytics response for a round trip from edge to cloud and back. An example is a car that senses a dangerous lateral rotation and uses analytics in the car s computer system to make the necessary braking, suspension, and speed adjustments to avoid an accident. Suppliers are also driving more access to edge computing features. Analytics in gateways have been steadily improving through improved hardware and application development toolsets. Computing power and memory in semiconductors and modules is increasing, allowing simple sensors to offer some level of edge computing, including advanced encryption and certificate security services. Network infrastructure and services providers are also seeing the opportunity for deploying analytics near the edge. CDN providers are contemplating IoT analytics services since they already offer server capacity close to users devices. Mobile network infrastructure OEMs are also a potential new entrant for analytics services from network base stations. The combined effect of both customer-driven and OEM-driven edge analytics is enablement of more futuristic IoT applications, such as delivery services using drones and autonomous transportation systems. Edge analytics will also greatly accelerate innovation in the data and analytics markets, helping IoT realize the potential of ambient intelligence, where products learn from their surroundings and get smarter over time. 8
DATA CENTER DRIVING THE IoT JOURNEY: 10 TRENDS TO WATCH LPWA TECHNOLOGIES DEMONSTRATE THE POTENTIAL FOR MASSIVE IOT CONNECTIONS AND BUSINESS MODEL DISRUPTION Low-Power Wide-Area (LPWA) technologies are the latest in a portfolio of connectivity technologies that include cellular, short-range wireless, satellite, and fixed line. They fill a gap in the market for application segments that only have batteries or energy capture for power, require low costs for connectivity hardware and services, and seek the simplicity of a wide area connection rather than architectures that use gateways and short-range wireless connections. All of these factors lower the total cost of ownership of IoT solutions. Proprietary LPWA technologies exist today but equivalent standardized versions will become available in 2017. IoT DATA EXCHANGE Besides opportunities in established markets, such as smart metering, smart home, and commercial building automation, LPWA technologies make IoT solutions affordable for nascent markets of asset tracking, agriculture, and environmental monitoring. In fact, IoT DATA the concept of the sharing EXCHANGE economy, where personal assets are loaned to individuals for a fee, is greatly accelerated with LPWA technologies. In sharing economies, anything from lawn mowers, to surfboards to the kid s saxophone can be tracked and monitored using LPWA technologies for much less cost than with existing wireless technologies. The sharing economy whether for consumer or enterprise products is an example of a new business model enhanced by LPWA technologies. But the significance of LPWA is that now the addressable market for connected products, and by extension innovative business models, is massively increased. Even more products can be servitized, which today is in the tens of billions but with LPWA could increase to the hundreds of billions and even trillions. LPWA technologies help facilitate not only massive numbers of connections, but also massive access to the new IoT business models that will shape the IoT market of the future. ENTERPRISE OUTSOURCING AND MVNOS FIND LIFE WITH IoT IoT is about connecting a vast array of things and machines. Each thing requires different technologies to both sense, connect, and transmit data. The data from each thing can be quite different and require a different set of analytics. The security requirements on things can range from very limited security to highly secure. The services required will be dependent on not only the thing but also the needs of the companies and partners involved in the connected thing s operation. IoT marketplaces provide a one-stop shop to build an IoT solution where the enterprise then typically runs and manages its connected equipment. But in some cases, it may make sense for an enterprise to outsource IoT activities, including building, running and managing a suite of connected things. The reason is that some enterprises may not want to invest the time and CAPEX to connect their products. Or IoT may be important but not the core part of their business. Regardless, the various connected things owned by businesses to facilitate their operations provides an opportunity for suppliers who can build, run and/or manage the entire IoT solution. 9
Outsourcing IoT services is not unusual but very much a nascent market. ATM management includes companies that provide connectivity, management, and/or cash-in-transit services for collecting and distributing money. Companies such as Ericsson are offering Smart Metering as a Service (SMaas), and companies like Philips are offering smart street lighting management services. Philips can be classified as a mobile virtual network operator (MVNO) if it buys the connectivity from mobile operators to enable the connection. IoT MVNOs will become a viable option for outsourcing services as wide area wireless technologies become more prominent in thing connectivity. As more things get connected, industries that deploy a vast array of IoT technologies for their operations will eventually look to suppliers to outsource not only their operation and management but complete cradle-to-grave services. No enterprise segments are excluded as potential customers but the early opportunities will be with cities, manufacturing, retail, and healthcare. SUMMARY Why do these trends matter? Because a smarter, more connected world is upon us with real benefits for both enterprise and consumer markets. More importantly, these trends show that enterprises have choices to address IoT solution enablement. From connectivity using LPWA technologies, to application enablement using one-stop shop suppliers, to analytics automation, to IoT marketplaces, innovative options are available. As well, by engaging today, enterprises will not miss future opportunities facilitated with AR solutions, and thing identity and data exchange services. As noted earlier from survey data, most companies are in the early stages of their IoT journey. Regardless of project stage, enterprises need to engage customers, partners, and IoT suppliers and embrace the opportunities and change imminent in a smarter, more connected world. 10
Published June 2016 2016 ABI Research Post Office Box 452 249 South Street Oyster Bay, New York 11771 USA Tel: +1 516-624-2500 Fax: +1 516-624-2501 The material contained herein is for the individual use of the purchasing Licensee and may not be distributed to any other person or entity by such Licensee including, without limitation, to persons within the same corporate or other entity as such Licensee, without the express written permission of Licensor. 2016 ABI Research ALL RIGHTS RESERVED. No part of this document may be reproduced, recorded, photocopied, entered into a spreadsheet or information storage and/or retrieval system of any kind by any means, electronic, mechanical, or otherwise without the expressed written permission of the publisher. Exceptions: Government data and other data obtained from public sources found in this report are not protected by copyright or intellectual property claims. The owners of this data may or may not be so noted where this data appears. Electronic intellectual property licenses are available for site use. Please call ABI Research to find out about a site license. DRIVING THE IoT JOURNEY: 10 TRENDS TO WATCH