Earlier this year, Gartner estimated that approximately 8.4 billion connected “things” will be in use in 2017, up 31% from last year and will surpass 20 billion by 2020 (other analysts estimate larger numbers). Most of those are consumer “things”, however Gartner estimates that business and industry will drive growth from 2018 onwards, as connectivity takes hold with higher volume, low cost devices. Hardware spending from consumer, business and industrial IoT segments is expected to reach US$ 3 trillion by 2020 from $1.4 trillion in 2016. Ubiquitous connectivity leads to gigantic data volumes, probably upending past growth projections, which saw extant data stocks doubling every two years, with human and machine data combined growing 10x faster than business data and machine data alone growing 50x faster.
Connectivity and Big Data together with quite recent advances in computation technology, in particular approaches to artificial intelligence centered on computational statistics and machine learning, probably constitute a paradigm shift encompassing not just business and industry, but all aspects of life -comparable perhaps to the advent of electrification in the late 19th century or the industrial revolution of the 1850s. Its consequences (including on employment) are still not knowable, but its evolution has been interesting -from slow and linear for almost 80 years to faster and exponential, almost explosive now.
Moore’s law from the 1960’s drove an exponential increase in computing power relative to cost, size and energy consumption. This, in turn, led to ubiquitous computing in the form of personal and laptop computers in the 80’s, the rise of the internet and the worldwide web in the ’90s and in the mid-00s to a pivotal product: the Smartphone (the iPhone is celebrating 10 years since its launch this year). Smartphones (to be understood as pocket sized devices with the power of supercomputers of the 1990s) ignited a number of reinforcing revolutions: i) in connectivity (speed, availability, reliability, cost) by driving demand – the utility of a smartphone is vastly enhanced as a communications and data transfer device; ii) in sensing and actuating technology and miniaturization (to MEMS -MicroElectroMechanical Systems) -smartphones required development of small, cheap, reliable sensors of all types in very large quantities, which drove down costs and improved functionality and energy efficiency; iii) in data volumes – smartphones themselves are a source of gigantic data streams in all forms, which set the stage for cloud computing, data science and machine learning and iv) the proliferation of platforms as commerce and interaction engines. The revolutions do not stop there of course: 3D Printing, Virtual and Augmented Reality and a “Cambrian explosion” in robotics (including autonomous vehicles of all types) are a direct continuation of previous developments. So is the Internet of Things (IoT) and its derivative for industrial settings -the Industrial Internet of Things (IIoT), sometimes also referred to as Industrie 4.0.
Internet of Things: “A global infrastructure for the information society, enabling advanced services by interconnecting (physical and virtual) things based on existing and evolving interoperable information and communication technologies” Internet of Things Global Standards Initiative
The IoT and the ensuing general digitization and automation of processes it makes possible will no doubt produce huge benefits in the aggregate. In a study from 2015, McKinsey Global Institute estimated an “Economic Value” of up to $11 trillion per year from “business” IoT by 2025 (for reference the total size of the global economy is approx. $80 trillion in nominal terms and $110 trillion in terms of Purchasing Power Parity). Such studies need to be treated with caution, as extrapolating from existing use cases is not a particularly valid way to forecast the future when dealing with exponential technologies and unknowable future applications. Nevertheless, it illustrates the potential. Economic value will be generated through more efficient resource allocation, higher productivity, better utilization of resources and reduction of waste (not calculated was the impact of new products/services on consumer utility). It will also drive the on-going “dematerialization” of the economy.
But while the impact on the economy as a whole is clear, the effect on individual companies is far less so. The IIoT is hugely disruptive in terms of its potential impact on business models, value and supply chains, operations, transactions and interactions. It removes many barriers to entry from previously “moated” markets and changes the rules of the game. Incumbents may lose market power. Costs may come down, but prices may come down even further. McKinsey estimates that 90% of the IoT economic value will be captured by customers/users, whether businesses, other organizations or consumers rather than vendors of IoT solutions and services and how the allocation of market between them will play out is, again, unknown at present.
As noted in a previous article on digitization more generally:
Most top executives expect substantial benefits [from digitization] -not only in productivity and operational efficiency, but also in terms of top- and bottom-line growth. That latter assertion however, slightly smacks of illusory superiority syndrome, where individuals (or companies) overestimate their own capabilities relative to others. For example, in driving skills surveys large majorities of drivers routinely rate themselves better than average.
For digitization to boost growth for everybody it must unleash a sustained demand and market expansion (volumes). But that is only a necessary not a sufficient condition. In addition, the expansion must be such that incumbents are somehow protected from new entrants and can resist digitization induced price pressures. This is not likely and so the actual results of industrial digitization will most probably be an assortment of winners and losers. Furthermore, realistically it is unclear whether industrial digitization will indeed drive revenue growth for incumbent vendors even on average.
This is because a primary expected result of digitization is dematerialization (achieving more or better outcomes with less resources) which almost always (in competitive markets) results in lower prices, in some cases close to zero prices. This leaves users/consumers with a surplus which they can either use to consume more of the same or something else (e.g. a new or different product or service) or save (which then translates into investment through the markets, probably in further digitization). It follows then that to grow the combined volume effect (up) must be greater than the price effect (down), while understanding that the most likely effect on volume is actually new products and services.
For example, it is unlikely that digitization of, say, wind turbines (in the sense that they become smart connected products), will drive a higher demand for wind turbines. Nor does it seem likely that digitization will drive an overall increased demand for wind power services, though it will cause a shift in the service-mix: smart services (e.g. data analytics based prognostics and optimization) will displace traditional services (e.g. repairs due to failures). As smart services are immaterial and have low marginal or variable cost (high fixed cost) their price will be low in competitive markets (i.e. in the absence of monopolistic structures – admittedly a caveat). The net effect will probably be a decrease in revenue for service vendors, eventhough the “value” of the service (e.g. as measured in production throughput (MWh) per $ value of input) will be much higher (increased user surplus). Of course, vendors do count on their ability to push through robust prices for smart services based on a perceived installed base advantage. However, smart services, in contrast to conventional services, are based on open infrastructural technologies to which anyone (eventually) has access, thus conveying limited sustainable competitive advantage. Price pressures therefore are likely to be considerable.
Strategically therefore for industrial companies, machine builders and manufacturers staying out of the IIoT/digitization bandwagon is simply not an option. Simply getting on board however does not guarantee results. The process needs to be carefully though through, requires good understanding of context and objectives as well as technology, competent management of change and, above all, a good understanding of how industry economics change with the Internet of Things and how to position in new ecosystems.
In its basic form, the Industrial IoT is about using insights from smart connected/networked products (devices, machines) to improve performance, offerings and decision making and enable new offerings or business models. To understand what it entails technically, a good starting point is the standard developed by the Industrial Internet Consortium as a three tier architecture, consisting of respectively edge, platform and enterprise tiers:
The edge tier collects data from the edge nodes (devices, machines/assets), using the proximity network. The architectural characteristics of this tier (breadth of distribution, location, governance scope, the nature of the proximity network) vary depending on the specific use cases. The platform tier i) receives, processes and forwards control commands from the enterprise tier to the edge tier and ii) consolidates processes and analyzes data flows from the edge tier. It provides management functions for devices and assets. It also offers general (i.e. non-domain specific) services such as data query and analytics. The enterprise tier implements domain-specific applications and provides interfaces to end-users including operation specialists and managers. It receives data flows from the edge and platform tiers and also issues control commands to both these tiers.
From a functional perspective, the edge tier implements most of the control function or “domain”; the platform tier most of the information and operations domains; the enterprise tier most of the application and business domains. Of course this is merely a standard. In real systems, the functional mapping depends on the specifics of the system use cases, requirements and the vendor’s architectural/technological approach. For example, some functions of the information domain may be implemented in or close to the edge tier, along with some application logic and rules to enable intelligent edge computing, something which can be important in some applications, particularly mission critical applications in process industries, aviation etc.
Currently many different vendors with a variety of backgrounds (software, automation, industrial -or combinations thereof) offer “industrialized” general (public) IoT or “tailor made” IIoT platforms to clients, usually as a service (PaaS). Their emphasis on various platform tiers usually reflects their backgrounds and target markets/applications. GE, for example, has made a company changing investment in digitization and IIoT through its Predix platform (accompanied by numerous complementary acquisitions). The Predix architecture adheres fairly closely to the Industrial Internet Consortium standard, but with particular emphasis on edge computing (GE mas many mission critical domains), while its central (cloud based) analytics approach utilizes a concept called the “Digital Twin” reflecting its background in large assets.
In Predix edge computing is handled through Predix Machine, a software to develop and deploy machine apps and monitor, collect and analyze data and manage machines locally without central intervention through the cloud when necessary. Predix Machine provides advanced “edge analytics” such as in-motion data analytics and machine learning capabilities with zero-touch application and analytics deployment at the edge. It runs on a variety of hardware platforms from sensors, controllers, gateways, to on-premise appliances and also provides security, authentication, and governance services for endpoint devices. This allows security profiles to be audited and managed centrally across devices, and that critical data is protected. Predix Edge Manager is used to manage the Predix Machine software running on the hardware.
The gateway acts as a smart conduit between the cloud and the machines – providing connectivity to assets via a variety of IT or OT protocols. By using existing controllers, industrial and commercial assets that previously operated stand-alone can be connected to the cloud for data collection and analytics. Leveraging low-cost intelligent sensors deployed on or near the assets allows data to be transmitted directly or through a gateway to Predix (cloud).
At the platform (cloud) tier, Predix provides a variety of asset management and analytics capabilities and functions, particularly around the “Digital Twin” concept, where by combining data from sensors, devices and other sources (including unstructured data) with analytics and operating models, a simulation of individual components or assets (or combinations, processes and plants) is created and is used as a standard for the monitoring, analysis and prediction of the behavior of the corresponding real assets in the field. As more digital twins run on the platform, the industrial learning system feeds back data to the individual digital twins, improving fidelity. By comparing behaviors of a variety of parameters, it’s possible to detect anomalies a long time before they become real problems and take appropriate action.
Predix is open to third party app developers (customers, independent vendors, other OEMs) and GE has partnered with major connectivity, infrastructure, analytics and other technology providers to accelerate development and roll out of solutions for customers. A good example of a solution using Microsoft’s Azure Business Intelligence analytics to analyze and determine a problem on wind turbines using Predix can be seen in the below video clip
At present numerous vendors have developed or are developing IIoT platforms which (with variations) adhere to the Industrial Internet Consortium standard. Some were developed as general or “public” platforms, but were successfully deployed in industrial settings. Others are based on more generic platform architectures and functionalities and have been “industrialized” (for example, issues of security. GE Predix is based on the open source Cloud Foundry). Furthermore, some vendors have approached the IoT from a specific capability background and extended functionalities from there, e.g. the cloud and computing infrastructure (Amazon AWS, Microsoft Azure), others from machine learning, data science and analytics (e.g. IBM Watson IoT), others from viewing the need to connect devices (Bosch, GE Predix) or to provide a bridge to ERP systems and other high level process management applications (Oracle, Salesforce, ThingWorx). The following table provides a non-exclusive list of major platforms currently available in the market.
|Name||PaaS Platform||Release Year|
|Amazon IoT||AWS IoT||2015|
|Ayla IoT||Ayla IoT||2014|
|Bosch IoT Suite 2.0||Bosch SI PaaS||2015|
|C3 IoT||C3 IoT Platform||2009|
|Connectthings IoT||Connectthings IoT||2007|
|GE Predix||Cloud Foundry||2015|
|Cisco Jasper||Jasper Control Center||2004|
|M2Mi||IBM Bluemix, AWS||2006|
|Azure IoT Suite||Microsoft Azure||2015|
|Oracle IoT Cloud Service||Oracle IoT Cloud||2016|
|PTC ThingWorx IoT Platform||ThingWorx||2015|
|Salesforce IoT Cloud||AWS||2017|
|IBM Watson IoT||IBM Bluemix, AWS||2016|
In addition, other vendors have developed IIoT platform offerings for different target markets, including Cumulocity, Kaa (open source), Losant, Zatar (Zebra Technologies), thethings.io, Exosite and Xively (Logmeln) -not to mention major players like Verizon, SAP (acquisitions of Plat.One and Zedem Technologies) or Google, whose platform is currently in beta mode.
A small number of automation, engineering or manufacturing OEMs have decided to invest in full blown IoT platforms, e.g. Siemens with Mindsphere. Most have decided to partner with major platform providers to develop specific or domain centered IIoT offerings and associated services (e.g. ABB’s Ability system is based on Microsoft’s Azure platform; Metso is co-developing IoT mining solutions with Rockwell Automation based, again, on Microsoft’s Azure platform and Komatsu is doing the same based on GE’s Predix ) or focus investment in particular capabilities such as general analytics while relying on other platforms for operating system functions, an example being Caterpillar’s investment in Uptake.
Based on a set of criteria, market intelligence firm Forrester Research has developed a ranking map for the main platform offerings currently in the market.
Nevertheless, the success factors for platforms center mainly around the capability of vendors to create an ecosystem of participants (customers and value added service providers) that creates network effects. How that may be accomplished in industrial settings and how OEMs (larger or smaller) can or should position themselves within IoT platforms, what their strategy and practical approach should be and what this means for services will be discussed in the second part of this article.
Si2 ON-Demand: Deep dive remote sessions with Experts, backed up by analyst and research resources to solve problems and get things done faster, with less mistakes, at lower cost and less risk! – A fresh new approach for Service Leaders and their teams. To find out more see our post on this blog or visit Si2 Partners
If you are a Service professional (manager, practitioner, consultant or academic) in an industrial setting join our group Service in Industry on Linkedin
We curate many Magazines on Flipboard on service markets and industries, service business and operations as well as service related technologies, the IIoT and innovation. You can follow us on Flipboard here . The content is crowdsourced. If you would like to be a co-curator, and share interesting articles with the community through Flipboard, please send us an email at email@example.com with the heading “Flipboard”.