De-mystifying Industrial IoT (IIOT) Platforms – A Personal Opinion

IOT followed by Industrial IOT (or IIOT) has become one of the commonly used jargons in the industry over the last few years. Seeing the upside potential of IIOT, software companies rushed into building IIOT platforms and assumed that they will be able to monetize them fairly quickly. As a result, a number of companies, small and large, built their IIOT Platforms. However, monetization was not as simple as they thought. That led to consolidation of platforms on one hand, and to creation of differentiated IIOT platforms on the other. Differentiation resulted from either adding additional horizontal (Industry-agnostic) functionality to the core platform, or adding vertical (Industry-specific) capability to the basic IIOT platforms. This article discusses the Technology trends in IIOT as seen by a practitioner in the field.

1.    Maturity of IIOT Platforms

As the business use cases for IIOT started to mature, and customers tried to decide on the IIOT Platform for their own use cases, it became apparent that almost none of the so-called IIOT Platforms could deliver the use cases reliably, scalably and within budget while delivering the desired business outcome. Customers, driven by specific business needs, started to delve deeper into the various Platforms available in the market to understand their pros and cons to narrow down on the Platform that satisfied their specific needs. What was interesting to observe is that different businesses ended up selecting different Platforms to suit their needs and there was no clear winner! The natural question was then why do different businesses find different Platforms best suited for their needs.

That led to the next level understanding of the core functionalities of an IIOT Platform so one could say Platform X is strong in component A and weak in component C while Platform Y is strong in component C and weak in Component B.

At the same time, every company that claimed to have its own IIOT Platform started to realize that although it claimed to have an IIOT Platform, it didn’t really have all the pieces/components of an industrial grade platform needed to succeed in the field. For example, there were companies with strength in machine to machine (M2M) communication who jumped in the fray only to realize that device management which is their core competence is but one of many components needed for a complete IIOT platform to succeed in the field. On the other hand, there were companies with strength in enterprise software who also jumped in only to realize that their ability to connect to Enterprise systems is also one of many components needed for an industrial grade IIOT platform. Similarly, companies with strength in data management or analytics did realize that they have a play in IIOT platform and started to claim that they have an IIOT platform to realize that they also fell short in some other dimensions.

All of the above experiences led the industry to really dig deeper and articulate what are the core components that would make an industrial grade IIOT platform and more importantly will help solve the real business problems at hand.

Here are the core components (or functions) of an industrial-grade IIOT Platform:

1.    Device Management

2.    Interfacing with Devices (Southbound Interface)

3.    Data Management

4.    Real-time Analytics

5.    Interfacing with Enterprise Applications (Northbound Interface)

6.    Security

Device Management is needed to configure, provision, monitor and troubleshoot devices remotely. Firmware Over The Air (FOTA) upgrade is a key functionality of device management layer of IIOT Platform.

Southbound interfaces are needed for connecting with the end devices. There are many variants, especially in an industrial setting, as there are multitude of protocols (such as, Canbus, Modbus, Backnet, etc.) on top of the lower layer communication protocols (such as, RS-485, RS-422, BLE, Zigbee, WiFi, LoRAWAN etc.). This necessitates the existence of what is known in the industry as Multi-Protocol Gateway (MPG) that can speak these protocols (usually a subset) on the southbound side and a standard protocol like MQTT or AMQP or HTTPS on the northbound side to the Cloud component of an IIOT Platform in most cases.

Data Management deals with the data from end systems streaming in real time to the Platform. The challenges are with the storage and processing of data in real time. How much to store in memory and how much to store in secondary storage? What kind of data bases to use? How can data accessibility be provided to IT and OT systems as well as to external systems. Flow of data and communication between various entities within the platform would also belong here.

Real-time Analytics deals with what is known as Complex Event Processing (CEP) which enables execution of business rules on the streaming data leading to the generation of notification for various events to alert the relevant business stakeholders. Batch analytics is also an integral part of IIOT systems but the statistical model development for machine learning etc. need not be included as part of the IIOT Platform. However, once a statistical or machine learning model is developed, it can be ingested into the IIOT Platform and executed in real time with the streaming data.

Northbound interfaces are necessary for building various business apps relevant to the industry in question. They help expose the core capability of the IIOT Platform, such as the generation of business insights from real time field data that can be leveraged to make the traditional enterprise applications smarter or more efficient.

Security has multiple dimensions in the context of IIOT Platforms. First, only authenticated devices are supposed to connect to the platform so that attacks cannot be launched from malicious devices. Second, data in motion as well as data at rest should be encrypted so no data can be compromised. When multiple parties use the same platform, there is also a need for role-based authorization. Privacy also plays an important role from compliance and audit perspective.

Once we have the above components of an IIOT Platform, we can start comparing various IIOT Platforms with respect to these functionalities. Broadly, there are two types of IIOT Platforms: (1) Industry-agnostic (or Horizontal) Platforms and (2) Industry-specific (Vertical) Platforms.

A.   Industry-agnostic IIoT Platforms

Most common IoT Platforms, such as, Azure IoT Platform and AWS IoT Platform focus on Data Management that capture and store streaming data reliably. They also introduce an asynchronous messaging channel for communication between components to make the data available in real time to Stream-processing Applications and Real-Time Analytics. Device management and northbound interfaces are also supported to a limited extent. However from southbound interface perspective, they support standardized protocols like MQTT, AMQP and HTTPS up to the IoT Edge Gateway or the end device with built-in IoT agents. Customers need to partner with OEMs for access to specialized industrial protocols like Modbus, Canbus etc.

Please note that the next section of this article is not aimed at doing an exhaustive comparison of IIOT platforms in the industry. There are several articles and analyst reports available just for that purpose. The objective here is to highlight the core strengths of various IIOT platforms and the reason behind that. When a given platform does not have core strength in one or more components of the IIOT platform that does not mean that the vendor does not have an IIOT platform for the clients. It just means that the vendor has tied up with other niche players to make up for the deficiency. The end goal for every vendor is to provide a comprehensive industrial-grade IIOT platform – which functionality comes from alliance partners varies from vendor to vendor. The cost, performance, scalability may vary. The final choice of the platform will be with the clients based on their specific needs of scalability, performance and of course, affordability.

With the above objective in mind, let us take a look at some of the more well-known IIOT Platforms in the industry and understand their strengths and weaknesses with respect to the above-mentioned core components of the Platform. Here are some of the representative platforms:

1.    Software AG’s Cumulocity Platform: it is one of the strongest when it comes to Device Management. It offers a comprehensive IoT device life cycle management capability, including device health and connectivity management, in addition to secure upgrades to the devices over the air. Software AG has to partner with OT vendors to provide integration with legacy systems and with system integrators for integration with enterprise applications.

2.    SAP Leonardo IoT Platform: as expected, SAP's experience in enterprise applications and middleware makes its Leonardo IoT Platform one of the strongest when it comes to northbound interfaces. However, when it comes to device management, SAP provides cloud-based device management, but does not natively offer device management for on premise deployments. SAP Leonardo IoT Platform relies on partners like Telit for on premise device management.

3.    Oracle’s IoT Platform: on similar lines as SAP Leonardo, Oracle's application-centric approach to IoT makes its Platform strong from enterprise integration or northbound interface perspective. It enables faster implementation, ease of management and out of-the-box connectivity to middleware and enterprise applications. However, it is inherently weak on the device management front. Oracle IoT Cloud Service provides some basic device management capabilities but when it comes to advanced services, Oracle mostly relies on OEM’s device management solutions and as a result, it requires customers to support multiple tools for device management, depending on desired functionality.

4.    IBM’s Watson IoT Platform: as expected, it excels in the Analytics component. Analytics capabilities in Watson IoT are easy to use and require less training than competitive platforms. But when it comes to southbound interfaces, it falls short of expectations. In order for IBM’s Watson IoT Platform to broaden its value potential to industrial enterprises, it requires customers to have multiple partnerships with OT vendors.

5.    PTC’s ThingWorx IoT Platform: PTC acquired multiple products (Axeda, ThingWorx, ColdLight and Kepware) exactly to build strength across all the layers of an IIoT platform. Specifically, Axeda is very strong on device management. It enables companies to establish secure connectivity, and remotely monitor, manage, and service a wide range of connected machines, sensors, and devices. Thingworx, on the other hand, provides a platform to quickly build industrial applications and helps maintains a marketplace that offers developers access to connectors and extensions to popular IT and OT apps and hardware. ColdLight provides PTC the strength in the Analytics layer via its big data machine learning and predictive analytics capabilities. Last but not the least, Kepware gives PTC strength in its southbound interfaces. Kepware provides communications connectivity to industrial automation environments and accelerates PTC’s entry into the factory setting and Industrial IoT (IIoT). These acquisitions have been consolidated into a single platform, ThingWorx and made PTC one of the strongest IIOT Platforms in the industry.

6.    Wipro Looking Glass: is a less publicized and less known IIOT platform in the industry but is one of the most comprehensive ones. It has excellent device management functionality in terms of configuration, provisioning and two-way communication with the device in a secure manner. Data management functionality is also strong as it supports an industrial strength time series database and an in-memory database for fast access to relevant data – whether raw or processed – needed during real-time processing. Real-time (or stream) analytics is available with full support for complex event processing and the ability to execute complex statistical models in real-time. Southbound interface is MQTT and HTTPS but with Wipro’s Multi-Protocol Gateway, a plethora of industrial protocols and/or connectivity to the end systems is supported. On the northbound side, standard connectors to ERP systems and other enterprise systems is also supported. Last but not the least, Wipro Looking Glass has a sophisticated asset modeling layer that enables various industrial applications related to asset management and digital twin while enabling sharing of the platform and data between multiple ecosystem partners with secure governance of access and privacy.

The above discussion is captured in the diagram below where the strengths of IoT Platforms are shown in the corresponding layers/components.

There are several new-age IIOT platforms that have been developed ground up just to address the challenges of real-world IIOT problems. Naturally, they have addressed all the above components of an industrial-grade IIOT platforms in a much more cleaner way compared to the above platforms who have addressed the challenges through partnerships. Examples of such platforms are Altizon’s Datonis, Ayla Network’s IOT Platform, C3 IoT Platform etc. There is a long list of such niche players. Then the question is, if they have all that are needed for an industrial-grade IIOT platform, why aren’t they then capturing the market. The answer to that is, IIOT is not just a technology play. It’s a business play and the platform is just an enabler for solving business problems. Clients want to solve their problems and/or generate additional revenue using IIOT and that requires building additional application software on top and these niche players do not have the scale or capability to do that. They need to partner with system integrators to be successful. Thus, it involves an interesting and complex dynamics to be successful in the world of IIOT driven business.

B.   Industry-specific IIoT Platforms

Industry leading OEMs, especially in Manufacturing industry, realized that they know more about their products and how they function much more than anyone else and hence they should be able to capitalize on that knowledge, especially when it comes to monitoring and controlling their equipment. These OEMs took a generic (industry-agnostic IIOT platform) and put a layer of intelligence on top, encapsulating their product-specific knowledge, leading to the development of Industry-specific IIOT platforms. Most, if not all of them, have started to create a marketplace of business apps on top of their own value-added layer of intelligence, creating a clear path towards monetization. Here are a few such platforms. Please note that this article is not meant to explore all industry-specific platforms in an exhaustive manner rather to give a flavor of what kind of values are added by the industry heavyweights in the field of IIOT.

1.    Siemens Mindsphere is positioned as an open, cloud-based development platform that offers an intuitive, straightforward environment that lets one build, scale and deploy IIOT based applications. Given that Siemens appliances are widely deployed in industrial manufacturing, healthcare, buildings, transmission drives, energy & utilities, Mindsphere, on one hand, has enabled Siemens to build its own value-added IIOT use cases, and on the other hand, has enabled Siemens to expose some of the core capabilities of its equipment for 3rd parties to build apps and make it available for others to consume. Mindsphere leverages AWS IOT Platform underneath with some other partner software components that are not publicly disclosed.

2.    Bosch IoT Suite as a Platform is a flexible IoT platform based on open standards and open source and provides all key middleware capabilities needed to build sophisticated IoT applications from top to bottom. Software developers can quickly set up prototype applications before implementing them, deploying them within minutes, and operating them at high availability. Exactly like Siemens, Bosch has also built its own apps and have exposed the capabilities of its equipment to 3rd parties to build apps on their equipment and make them available on the app marketplace. Bosch IoT Suite also leverages AWS IOT Platform underneath with some other partner software components that are not publicly disclosed.

3.    Honeywell Sentience IoT Platform based on Azure IoT Hub, delivers robust and secure big data capabilities for all of Honeywell’s connected solutions and is positioned as an overarching software platform and ecosystem. Engineering, instrumentation, aerospace and defense are their main areas of strength and they have built their Sentience Platform exactly to monetize on these capabilities via an app marketplace.

4.    Schneider Electric’s EcoStrXure based on Azure IoT Hub, is positioned as a plug-and-play, open architecture IIOT platform that delivers end-to-end solutions in six domains of Schneider Electric’s (SE’s) expertise — power, IT, building, machine, plant and grid. With an open architecture, developers, partners and customers can securely interact with and share data via SDK and API. EcoStruXure is very well-developed when it comes to leveraging SE’s products in the above six areas. App marketplace not only has SE’s own apps say in the area of power and building management, but also has some 3rd party apps built on SE equipment, especially in areas where SE does not directly want to play but want to increase usage/sale of their equipment.

5.    ABB Ability is a comprehensive and secure application platform to run, scale, and extend digital industrial solutions. The platform delivers shared capabilities that industrial applications require: asset connectivity, edge technologies, analytics and machine learning, big data processing, and asset-centric digital twins. ABB Ability has added a value-added layer on top of basic Azure IoT Hub enabling functionalities in areas of their strength, such as, power & electric, industrial automation, transmission drives etc.

One way to visualize these Industry-specific IIOT Platforms is shown below:

Desired state of IIOT App marketplace should be Industry and Corporation agnostic as shown below:

Note that the IIOT App Marketplace available today is specific to an industry-specific IIOT platform. For example, Schneider Electric offers an App marketplace for EcostruXure just as Siemens offers an App marketplace for Mindsphere. At first blush, it seems that the platform is bringing the entire ecosystem together for building complex IIOT apps but on a closer look, it is clear that unless you are an ecosystem partner of Schneider Electric and interface explicitly with EcoStruXure, you won’t be able to build an ABB Ability application that works seamlessly with Schneider Electric’s appliances for example. The same is true with Siemens and other vendors. Holistically however that should not be the case. 3rd parties should be able to build applications spanning multi-vendor equipment as it is hardly the case that a client uses equipment from just a single vendor. If that is the case, there needs to be a Platform-level support that would allow ecosystem orchestration across multiple vendors leading to multi-vendor IIOT apps. Wipro Looking Glass, by virtue of its sophisticated Asset Modelling framework, is in the best position in the industry to enable that as far as the author knows. That is why the diagram above shows Wipro Looking Glass at the topmost layer of the platform.

2.    Edge Networks

In addition to maturity of IIOT Platforms, another trend is the development of the Edge Gateway in the end-to-end IIoT architecture. There are several reasons for the Edge Gateway to become a trend in the end-to-end architecture:

1.    Many end systems are intermittently connected to the Internet and hence should be able to function even when they are not connected to the Internet. For example, a ship or an aircraft or a far-flung factory may not have continuous connectivity to cloud. However, the system should not cease to operate.

2.    Need for fast response may not have the luxury of sending data to the cloud, making decision there and sending the action back to the premise. For example, in an electrical sub-system, when things go wrong, local circuit breaker needs to be tripped quickly.

Here are some of the Edge systems in the industry and they fit in Stage 3 (Edge IT) of the 4-Stage IoT Solutions Architecture shown below:

1.    HPE Edgeline converges IT (compute, network, memory) and OT into one device to deliver data center-level computing power to edge locations enabling analytics at the edge, smart computation, aggregation and filtering of IoT data at the edge. This helps improves latency, bandwidth, cost, compliance, and security.

2.    Dell Edge Gateways for IoT – These are industrial-grade, intelligent devices that help capture and analyze data from sensors, endpoints and other devices (Stages 1 & 2 in the diagram above) to send meaningful data to the cloud, control center or data center. Dell Edge Gateways are capable of performing analytics at the edge of the network, thereby enabling smart IOT applications.

3.    Cisco IoT Networking is based on the concept of fog computing that shows how the IoT and cloud computing bring connected devices into distributed computing, right at the source. Cisco has a diverse range of IoT compatible routing and connectivity devices that use many different wired and wireless connecting technologies to integrate disparate devices into coherent networks. Using these products, Cisco offers simple, automated, intelligent, and secure IoT connectivity from edge to cloud. Cisco also offers an Industrial Compute Gateway with built-in security and manageability at scale, enabling faster decision making at the edge by processing and normalizing data from various sensors. Cisco fog computing is a core enabler for distributed intelligence making smart cities a reality.

4.    Wipro Multi-Protocol Gateway (W-MPG) is an IoT Edge Gateway that connects to the Wipro Looking Glass IoT Platform on the northbound side using MQTT and/or HTTPS and to the devices on the southbound side using multiple protocols like Modbus, Canbus, etc. and wireless protocols like Zigbee, BLE, Wi-Fi, LoRaWAN as shown below. Over The Air (OTA) updates, remote configuration and remote diagnostics are supported by W-MPG. W-MPG also supports edge computation, and thereby enables applications to be run at the edge doing lightweight analytics, filtration and aggregation of data at the edge before uploading to Looking Glass making the end-to-end IIOT system scalable, responsive and secure. W-MPG also enables several value-added intelligent operations depending on context.

3.    Extreme low-cost, low-energy sensors/devices in a wireless local area network

In the world of manufacturing, there is a need to accurately monitor machines and equipment to ensure optimal operation of the plant. Many different technologies can be used to achieve this objective but what separates them is the cost to implement and how much energy and resources are wasted in the process.

The compromise is between getting the job done satisfactorily, while being as cost effective as possible. There are several local area network technologies that are very robust, and at the same time, are relatively low cost and have long ranges. However, the problem is the high cost of installation. Modern wireless technologies provide a much more flexible way of connection, reducing the cost of installation, while delivering the same high standard of data delivery. Wireless technologies, such as, ZigBee and IEEE 802.15.4, can be very expensive and use a lot of power even though they provide acceptable performance.

Bluetooth, on the other hand, has quickly become a most promising alternative to wired and other wireless solutions and building upon the original Bluetooth standard, is a new version called Bluetooth Low Energy (or BLE). The technology builds on classic Bluetooth and is in many ways, very similar and it therefore has all the advantages normal Bluetooth has over other technologies, while having a few extra ones when it comes to efficiency and therefore the cost of the service as a whole. One of the advantages that BLE, like classic Bluetooth still has is the ability to hop between frequencies and this makes the technology very robust. Wireless sensors have been developed with Bluetooth Low Energy that help check the lifecycle of machines. This is more reliable than other sensors as it has better battery life, saving energy when the Bluetooth is not necessarily needed and autonomously coming back on when it is needed. Here are some real-world products that use BLE effectively:

1.    A classic example of low-cost low-energy sensor is a product called BluVib from Sensor-Works. It is a wireless machine condition sensor, measuring both vibration and temperature data, doing data collection, data analysis and low power wireless communication. Data are transmitted using the latest Bluetooth 4.0 Smart Low Power standard in order to optimize and extend battery life.

2.    Another example is the mesh of low-cost low-energy intelligent IoT sensors creating what is called a Sensor Fabric from a company called Cloudleaf. The Sensor Fabric shown below consisting of sensors and gateways is used for continuous tracking of raw materials from suppliers to the trucks/vehicles carrying them to the factory of the manufacturing company to the finished goods delivered via trucks/trains/ships/vehicles to the distribution centers and/or to the end customers.

Thus, there is a trend in developing extreme low-cost, low-energy sensors/devices that can function for a long time without a change of battery. These devices are useful in track & trace of raw materials and/or finished goods through the supply chain.

Summary

After a lot of soul-searching, the industry is realizing that there are a variety of reasons because of which it is taking so long for IIOT applications to be deployed in practice. One of the main reasons, argued in this article, is the lack of maturity of an industrial-grade IOT platform that provides the desired performance, scalability, security and responsiveness at the desired price point delivering the business use cases. However, these gaps are quickly being addressed by larger software vendors with partnerships and alliances with niche players specializing in certain core functions. The other way this need is being addressed is through brand-new start-ups who have built their IIOT Platforms ground up just to address the exact needs of IIOT applications, and they partnering with system integrators to build business use cases for industrial customers. While the above is a horizontal play, there is a vertical play as well. The manufacturing heavy-weights, such as Siemens, ABB, Honeywell, Schneider Electric, Bosch etc. are positioning themselves as software companies as opposed to traditional industrial companies and adding their core capabilities as a software value-add on top of core IIOT platforms and driving their customer base to the new world of IIOT-enabled applications and services. That is a great strategy for them to remain relevant to their customers in the digital age, and not be disrupted by technology companies. The bottom-line is IIOT is still in its infancy when it comes to real-world deployment and adoption. However, with the maturity of the platforms and apps and the plummeting cost of services, it is just a matter of time before we see IIOT cross the chasm and become mainstream.

Disclaimer: the above article is purely my personal opinion and does not in any way reflect the opinion or perception of Wipro Ltd.