Blueprints and feasibility studies for Enterprise IoT (Part Two of Three) 1 Executive Summary The Internet of Things provides a host of opportunities for enterprises to design, develop and launch smart connected devices, integrated with enterprise systems through new cross-platform applications and enhanced with services based on advanced analytics. As a result, enterprises have the additional opportunities to develop new business models, explore new customer engagement structures and potentially develop new revenue streams. The broad-ranging impact of Enterprise IoT is changing the rules of business. To realize these service opportunities, enterprises will need to identify, select and prioritize machine-tomachine (M2M) and Internet of Things (IoT) related opportunities and technologies. As part of this process, executives will need to consider a number of crucial questions around the design of an Enterprise IoT technology architecture. These architectures are substantially more complex than traditional IT implementations, and with comparatively greater degrees of openness envisaged. Three important pillars in a successful IoT design are scalability, agility and flexibility. Navigating through this complexity will be a challenge for any enterprise, and whether guided by a build or buy approach, enterprise executives will need to approach IoT architectures more as strategic blueprints for technology futures than operational here and now technologies. page 1
2 Contents 1 Executive Summary 1 2 Contents 2 3 Scope 3 4 What are the technological building blocks in an Enterprise IoT architecture? 4 5 Importance of scalability, agility and flexibility in IoT 7 6 Preparing a feasibility study for IoT opportunities 8 7 Conclusions & recommendations 11 8 About ThingWorx 12 9 About Machina 13 page 2
3 Scope This three-part White Paper series looks at Enterprise IoT from the viewpoint of enterprises, and act as a blueprint for enterprise executives to effectively address and make the most of this disruptive business and technology force. In Part One of this series, entitled A new agenda item for enterprise executives: Enterprise IoT, we had a close look at how executives will be able to identify, prioritize and initiate new Enterprise IoT projects. In this, the second paper, we begin to examine the more operational elements of Enterprise IoT, looking to develop a high level understanding of the different technology domains and components in an IoT architecture, and highlight the important factors that executives should take into consideration when starting on a blueprint for Enterprise IoT. This paper also provides an initial perspective on enterprises preparing a high level feasibility study before finalizing which applications to pursue. The aim of this Paper is to enable enterprise executives to recognize and appreciate the complexity and wide range of technical components in an Enterprise IoT solution, and act as a final short-listing tool before starting the business case phase. In Part Three of the White Paper series, we will explore the partnerships and collaborations enterprises will need to develop in this new Enterprise IoT environment, and start to touch on some of the key elements in an IoT business case for enterprises. page 3
4 What are the technological building blocks in an Enterprise IoT architecture? IoT solutions will depend on fairly complex architectures. Smart connected devices will be connected to local and/or wide area networks to allow for communications to take place between the devices and the applications. In the service layer domain, enterprise systems domain and data and applications domain, data will be managed, processed, stored and transformed into actionable insights. For many existing M2M and IoT architectures, the scope and functionality within each of these domains will have been tightly defined for the specific attributes of the application, and any associated regulatory, risk management, monitoring and reporting reasons. For the emerging IoT architectures, each of these domains will need to reflect the significant attributes of scalability, agility and flexibility. As illustrated in Figure 1, these six major domains make up the architecture of a typical IoT solution, and in the following sections, we explore each of these attributes. Figure 1: Major domains in an IoT solution architecture [Source: Machina, 2015] On-premise or Cloud based solution Software as a Service / Platform as a Service Device Domain (devices, sensors and modules) Wide Area Network Domain Service Layer Domain (ncl, M2M & IoT platforms) Local Network Domain (wireless sensor + router, hub or aggregator) Data & Application Domain Enterprise Systems Domain page 4
Device domain Device domains will remain the frontline in managing the increasing number of connected devices. In smart cities, for instance, such connected devices will include traffic lights, street lights, environmental monitoring devices, public transport signs, display signs, and occupancy sensors for parking spaces. In homes, audio-visual systems, lighting systems, security systems and home power management and temperature systems are becoming connected. Within the device domain, this wide variety of devices are connected in a number of different ways. Some use private networks, either mesh or hub-andspoke, while others connect via a public wide area network connection, for example a cellular network, to an external service domain. It is the continuous growth in numbers of different sensors, devices, and modules that will deliver the increasing complexity in protocols, data and overall architectural management for IoT. As enterprises explore the opportunities that arise from access to rich IoT data, and continue to implement more and more devices, platforms will need to be able to rapidly scale to meet the requirements of all these devices. The device domain also presents a new dilemma for enterprises. In the case of many earlier investments and technology deployments, products had short lifespans. In IoT, particularly when assessing such applications as remote oil and gas solutions, smart meters, renewable energy devices and intelligent building solutions, the lifecycle of such devices may reach 15 to 20 years. In terms of managing a device domain, enterprises will need to balance the immediate and continuous development of deployed assets as well as the longer term management of these assets, in many cases, pushing enterprises to design completely new service management models for their IoT solutions. Local and Wide Area Network domains The network domains will be the first areas where scalability, agility and flexibility will be tested. Connected devices will be far from homogenous, delivering an increasing range of protocols and diversity of data, and establishing for many applications, different sets of communications commands and patterns. Numerous network technologies are used in supporting the IoT depending on the demands of the device, its location, the amount of traffic and many other parameters. Options include short range technologies such as ZigBee, Bluetooth and WiFi, and wide area cellular technologies. With the emergence of IoT, new technologies, such as low power wide area (LPWA) networks like LoRa and SigFox, have also started to see larger deployments. Enterprises will need to explore each option with the planned applications. Common to the challenges to all the network technologies and domains will be the volumes, latency and bandwidth demands and ultimately resilience and security factors around data and applications. These will require enterprises to constantly ensure that not only are the suitable and cost efficient connectivity options identified for the appropriate applications but that these networks are properly monitored and managed across changing circumstances such as technology sunsets and upgrades, application suitability and proper resilience. Service layer and enterprise systems domains M2M and IoT platforms inhabit the service layer domain. These platforms act as the central junction, or middleware, for all device data communications and connect applications and enterprise systems to the remote connected devices. Through integration frameworks and APIs, levels of interoperability between platforms and various enterprise systems such as ERP, CRM, and financial systems are enabled. It should also be mentioned that integrations to enterprise systems include workflow integrations as well as integrations with mobile handsets and handheld data devices. Enterprises will need to ensure that platforms in the service layer domain are scalable, agile and flexible. As one example of achieving these attributes, these services may be delivered as a cloud based service, leveraging such capabilities as rapid scalability. Platforms in the service layer domain have also become important ways to remove friction in the management of applications and data. By designing IoT architectures with separate data and applications domains, platform providers such as ThingWorx have been able provide enterprises with tools which allow for flexible, agile and rapid application development and advanced usage of the data as generated by the rich device estate. Having started down the IoT path, enterprises will be quick to explore and discover additional opportunities from new connected devices, and the aggregation page 5
or development of new applications. With more and more applications being developed and ready-off-theshelf mash-ups, and advanced data analytics becoming more available, application development with platforms such as ThingWorx have significantly reduced application development times. This inherently means a quicker time to market, lower development costs and potentially new and innovative IoT services. Data and applications domain In M2M solutions, the management of data and applications was an integral part of the platform. These were tightly programmed in line with all others functions. In IoT, the separation of data and application within a separate domain highlights the significant changes happening within the architectures. The integrated and tightly knitted programming will have benefits such as robust security features. However, in an IoT environment where requirements will be continuously developing, and new applications will be launched according to business needs, such application development approaches limit the agility and flexibility required by enterprises. Unless managed as a completely separate domain, changes will become highly cumbersome, time-consuming and ultimately costly. A further development has been the disassociation of data from the architectures. In IoT architectures, data continues to be managed and processed for primary applications (the applications for which the architectures were originally designed). However, enterprises have recognized that data sets from one IoT architecture may in fact add value to other IoT architectures and applications. By disassociating data from IoT architectures, or making it accessible for other applications, a range of opportunities has opened up to design and develop new services based on combinations of data sets from different sources, potentially delivering unparalleled insights. Blueprints for IoT architectures are complex and involve substantially more expansive ecosystems of technologies and service providers. Within each domain, a range of technology options is available, and while enterprises will not necessarily need to navigate through the complexities of architectures for each application, enterprises will certainly be advised to explore those technologies and enablement tools which remove as much friction as possible in launching new services. page 6
5 Importance of scalability, agility and flexibility in IoT There are two sides to the arguments for scalability, agility and flexibility. First, enterprises responding to the expanding environment of connected devices will require platforms and architectures that are scalable, agile and flexible to be able to keep up with the changes in the technology landscape. These architectures may be to launch specific IoT applications, or connect with other architectures to leverage the data produced for existing services. The second, and more important, reason for these platforms and architectures to be scalable, agile and flexible is to enable enterprises to design and develop portfolios of new and innovative services. For many enterprises, one of the major challenges presented by IoT will be the constant requirement to explore new services and be aware of what new applications are emerging on the market. In either case, these architectures will need to be able to manage and scale not at an incremental pace but at accelerated rates as the world of things, applications and data begins to connect to more and more application opportunities. In this environment, enterprises will be responding to growing markets, and to remain competitive, these systems should not become the major contributors to time delays or costs. But why is scalability, agility and flexibility so important? Take the area of data management as one example. The diversity of data managed from structured transaction, machine or sensor data, to social media data and free form text, and to the extremes of unstructured data images, audio and video will require architectures to be as agile and flexible as possible. While managing and processing structured data will be possible with, for example, column-based databases with fixed schema, processing unstructured data will require a more open and flexible approach as provided by NoSQL databases. 1 Taking one step back, the challenge already begins with scalability. Here, new distributed storage methods as managed by, for example, DataStax and Cassandra have been introduced to deal with the tsunami of data that needs to be ingested and stored in data management systems. Taking one step forwards, querying these diverse types of data will also require highly agile and flexible query methods, allowing enterprises to add, amend and initiate new queries against new data sets on the fly (without having to re-establish completely new schema, data models, etc.). In the Internet of Things, scalability, agility and flexibility are set to become key attributes in future architectures. In Part One of the White Paper series, we looked at the ways to identify and prioritize the opportunities in the emerging IoT market. In Part Two we have shared, at a high level, an overview of the technologies and domains that make up an IoT architecture. With these two steps now completed, enterprise will be able to move forwards to the next decision-making step: a high level feasibility study. There are of course many ways to complete a feasibility study, and in the next section, Machina will share one of the ways enterprises may wish to approach this topic. Ultimately a more detailed business case will need to be prepared for the specific applications and final architecture. 1 For more information, read Machina research research Note on Why NoSQL databases are needed for the Internet of Things, published in April 2014 page 7
6 Preparing a feasibility study for IoT opportunities IoT applications such as tracking, remote diagnostics and maintenance in manufacturing and processing industries or for vehicles and machinery in agriculture as two examples will be composed of a wide range of characteristics, and one simple yet efficient way of starting the feasibility study is to outline the main characteristics of the application. For instance whether or not the application with be locally, nationally, regionally or globally implemented will have an impact on, for example, the regulatory assessment, or the integration complexity anticipated. In the cases of heavy industry machinery, IoT solutions here tend to be regionally designed and implemented, and the complexity may in fact vary from simple capture and forwarding of data to a more complete integration of the IoT solutions into workflow situations. A high level feasibility study can be based around eight attributes of any IoT application or use case: general market opportunity, revenue potential, strategic fit with the goals of the enterprise, time to market incl. channel to market, strategic importance for the enterprise, cost to implement, margin potential, and regulatory environment. For each of these attributes, a score of 1 to 10 should be assigned by the enterprise during the assessment process. An individual low result (from 1 to 3) will indicate less of an opportunity, fit or importance for the enterprise. An individual high result (from 8 to 10) will indicate a more significant opportunity, fit or importance for the enterprise, and certainly worth exploring in more detail. The attributes have been equally weighted in the proposed model but enterprises will be able to further develop the model by weighting individual elements. Figure 2: Illustrative set of high level attributes for a feasibility study [Source: Machina, 2015] Attribute Market opportunity Description To what extent is there a visible demand or need for the application either as a standalone solution or as an integrated option that can be exploited by the enterprise as competitors have not started to address the market. If demand or need is high, a score of 8 to 10 should be assigned. Revenue potential Strategic fit (with goals of enterprise) Time to market Relative extent to which an enterprise has the opportunity to generate revenues from the application. To what extent does the application align itself with the strategic aims of the enterprise. The expected time frame within which the application would be available for market launch. This assessment should include both technology factors in commercially launching the application as well as assessing the channels to market challenge at a high level. If the potential is high, a score of 8 to 10 should be assigned. If the alignment is high, a score of 8 to 10 should be assigned. If the time to market is relative short (as compared to other new product launches within the enterprises for core products and services), a score of 8 to 10 should be assigned. Strategic importance for the enterprise To what extent does the application present itself as a strategically important solution for the entire business of the enterprise, particularly given competitor initiatives. If the alignment is high, a score of 8 to 10 should be assigned. page 8
Cost to implement On a relative scale, what are the costs to implement the application. If anticipated costs are low or impact the implementation of applications in a minimal way, a score of 8 to 10 should be assigned. Margin potential Regulatory environment Relative extent to which the enterprise has the opportunity to generate suitable margin potential from the application. Certain applications will be subject to regulatory constraints which may impact the ease of implementation or adoption. If the potential is high, a score of 8 to 10 should be assigned. If few constraints exist or impact the launch of applications in a very minimal way, a score of 8 to 10 should be assigned. A fairly effective way of presenting these result is with a Radar chart (also known as a Spider chart) as illustrated in Figure 3. This method is suited for the individual assessment of potential IoT applications for the enterprise, and provides a quick and simple overview of the strengths and weaknesses in terms of attributes of an application. Individual attribute scores summarized also provides an application hash total which would allow for a degree of comparison between different applications. If weighting between the attributes is included, then an overall application score would be more representative, allowing for also for comparative assessments between all applications. Figure 3: Sample illustration of Feasibility Study assessment per application [Source: Machina, 2015] Regulatory environment 10 Regulatory environment Margin potential 8 6 4 2 0 revenue potential Strategic fit (with goals of Cost to implement Time to market Strategic importance for the enterprise page 9
Figure 3 is an example of an agricultural vehicle tracking IoT application being considered by a farm tractor manufacturer. The application will allow for the manufacturer to offer such additional services as predictive maintenance and warranty management of the tractor as well as immediate location services, on-board diagnostics, and additional value added services based on the data from the tractor and other IoT applications. The application will be implemented in all produced farm equipment, and deliver significant benefits to farmers and farming communities where sharing of agricultural equipment is crucial. In this assessment, illustrated in Figure 3, the market opportunity and strategic fit with the goals of the tractor manufacturer are significant. The number of farming communities are substantial and the need for efficiency improvements is an ongoing challenge for them. For the manufacturer, extending their business to include new services will definitely be a welcome fit. Implementing such a solution ( cost to implement ) is relatively low due to falling module prices as well as optional connectivity technologies and the margin potential from the various value added services ( margin potential ) will be very high given the potential ROI. However, the industry itself does show a slower adoption rate as indicated through the combined story of overall revenue potential, time to market, and the strategic importance for the enterprise. Adoption is happening but not at an accelerated pace. Connecting tractors will present substantial opportunities for tractor manufacturers but as recognized through this assessment method, it is not an immediate revenue solution for manufacturers and should be recognized as an important and strategic way forward to building longer term and direct relations with farmers and farming communities through the services. page 10
7 Conclusions & recommendations Preparing a blueprint for an IoT architecture involves having a good understanding of the technologies and domain options available to an enterprise, and having an overview of what some of the key considerations will be for the architecture. Enterprises are entering new fields with IoT architectures, and, as presented in the next White Paper of the series, working in partnership and collaboration with other service providers is a characteristic of this market. Machina would recommend to all enterprise executives to start with a blueprint of the Enterprise IoT architecture by: n Identifying at a high level the major components (domains) in an IoT architecture. Enterprises should identify each of these domains or building blocks in an IoT architecture, and explore the benefits (and constraints) from a significantly more modular approach where for example, platforms will be deviceor connectivity-agnostic. n Aligning IoT architectures with the requirements of the different applications. Different M2M and IoT applications will have different characteristics and requirements. Enterprises starting down the IoT road should assess at a high level the common features of envisaged applications, and design their blueprint in accordance with these requirements. Machina has designed a tool, the DNA of M2M, 2 which will assist many enterprises with this process. n Narrowing the process of application selection and prioritization. Understanding, even at a high level, the technology components of an IoT architecture assists enterprises in preparing a blueprint for the IoT architecture, and moving towards a feasibility study. n Preparing for the next set of applications. Enterprises are complex structures. They will be aligned with a specific industry vertical such as retail, healthcare, or intelligent buildings. Looking deeper into the structure of the enterprise, it is made up of multiple departments and functions such as sales, operations and manufacturing, logistics, administration, finance, customer services and so on. Each of these departments will in fact identify opportunities and potential applications in IoT. These IoT solutions may be for external or internal purposes, and will contribute in equally important ways to the generation of data. Ultimately, IoT architectures will need to support all this wide variety of applications, which returns us to the importance of scalability, agility and flexibility in IoT In Part Three of this White Paper series, we will continue to explore how partnerships and collaborations will be crucial for enterprises to address the full potential of IoT revenue opportunities in this constantly changing market of Enterprise IoT, and start to outline some of the elements in an IoT business case. 2 The DNA of M2M is available as an online tool for enterprise and service providers to explore the make-up of M2M and IoT applications. The tool identifies 27 characteristics such as security, billing, device integration, geographical coverage, analytics, access to power, and application complexity, adn assist companies to understand the requirements of each application. page 11
About ThingWorx ThingWorx, a PTC business, provides the first platform designed to efficiently build and run the applications of today s connected world. ThingWorx s model-based design and search-based intelligence simplifies application development efforts by minimising cost, and risk while accelerating time to value. The ThingWorx platform combines the key functionality of Web 2.0, search, and social collaboration, and applies it to the world of things, including connected products, machines, sensors, systems, and industrial equipment. Businesses use the ThingWorx platform to rapidly deliver innovative applications and connected solutions across markets ranging from manufacturing, energy, and food, to Machine-to-Machine (M2M) remote monitoring and service, as well as in emerging Internet of Things applications, including smart cities, smart grid, agriculture, and transportation. For more information, please visit our website at www.thingworx.com and our blog at www.thingworx.com/blog or follow us on Twitter at @ThingWorx. page 12
About Machina Machina is the world s leading provider of market intelligence and strategic insight on the rapidly emerging Machine-to-Machine (M2M), Internet of Things and Big Data opportunities. We provide market intelligence and strategic insight to help our clients maximize opportunities from these rapidly emerging markets. If your company is a mobile network operator, device vendor, infrastructure vendor, service provider or potential end user in the M2M, IoT, or Big Data space, we can help. We work in two ways: n Our Advisory Service consists of a set of Streams covering all aspects of M2M and IoT. Subscriptions to these multi-client services comprise Reports, Notes, Forecasts, Strategy Briefings and Analyst Enquiry. n Our Custom and Consulting team is available to meet your specific research requirements. This might include business case analysis, go-to-market strategies, sales support or marketing/ white papers. Machina s Advisory Service provides comprehensive support for any organisation interested in the Internet of Things (IoT) or Machine-to-Machine (M2M) market opportunity. The Advisory Service consists of thirteen Streams (as illustrated in the graphic below), each focused on a different aspect of IoT or M2M. They each provide a mixture of quantitative and qualitative research targeted at that specific sector and supported by leading industry analysts. Machina s analysts also have a wealth of experience in client-specific consultancy and custom research. Typical work for clients may involve custom market sizing, competitor benchmarking, advice on market entry strategy, sales support, marketing/ promotional activity, and white papers. For more information, refer to our website at https://machinaresearch.com, or email us at enquiries@machinaresearch.com. Advisory Service Streams [Source: Machina, 2014] Smarter Cars Future Wellness Smart Cities Enterprise IoT Connected Car Connected Health Connected Cities Connected Industry Connected Living & Working M2M Forecast Database IoT Strategies M2M Strategies M2M & IoT Regulation page 13