The supplier landscape for ML software and tools that can be applied to IoT datasets is complex, is growing and will coalesce around different consumption modalities.

To simplify the landscape, we’ve focused on vendors that can provide ‘one-stop’ shopping in that they have AI/ML capabilities along with traditional IoT platform capabilities (device communication, device management, data management, data visualization, etc.).

Basically, there are DIY models of consumption and engagement and there are application-centric ones. For the latter, the application comes with AI capability embedded, requiring no action from the end user. In cloud terms, it would be the difference between PaaS and SaaS.

Some vendors will offer both options. Many ISVs prepackage ML as part of an IoT-enabled application experience requiring almost zero expertise or fundamental understanding on the part of customers. Many suppliers will offer a variety of consumption models; the decision will ride on bespoke situations across workloads, customer skills and complexity.

This report gives you high-level overview of the industry’s top players in the ISV space for:

  • IoT and AI Infrastructure Platforms
  • IoT + OT Platforms
  • Application-Centric IT



IoT and AI Infrastructure Platforms




Alibaba wraps its cloud computing, AI and IoT efforts together via a narrative where cloud computing serves as the system’s heart, AI is the brain and IoT forms the nerves.

Alibaba Cloud offers both an IoT platform (Alibaba Cloud IoT Platform) and ML platform (PAI 2.0). Alibaba also packages up its IoT, Cloud and AI capabilities for specific industries such as healthcare (ET Medical Brain) and manufacturing (ET Industrial Brain).

Intel and Alibaba announced a partnership in September 2018 around a concept called the ‘Joint Edge Computing Platform,’ which features Alibaba Link Edge Solution and Intel’s OpenVino AI-based surveillance toolkit along with its processors.




AWS, the leader in cloud infrastructure services, offers a managed service for developers and data scientists to create ML models and generate predictions using data stored in AWS cloud. It’s extremely popular with developers, both for the pace of its innovation and for its budget-friendly, pay-per-use prices. The service, called Amazon SageMaker, was launched in late 2017; in late 2018, the company announced that more than 10,000 customers were using the service.

AWS launched its IoT platform in 2015 (AWS IoT Platform) and has been adding capabilities such as enhanced security (AWS IoT Device Defender) and analytics (AWS IoT Analytics).

At re:Invent 2018, AWS introduced additional services to make it easier to collect, monitor and process industrial IoT data (AWS IoT SiteWise, AWS IoT Events) and a new low-code, drag-and-drop app development service (AWS IoT Things Graph) to make it easier to build IoT applications. The core platform includes device connectivity (MQTT, HTTP and WebSockets supported), a device gateway, a message broker, authentication and authorization of these connected devices, a device registry and device shadow (basic operational state twin), as well as a rules engine to parse inbound messages and route them to the appropriate AWS or third-party service.

For edge computing use cases, the company offers AWS IoT Greengrass, which can support local machine learning workloads via AWS Greengrass Inference. The new capability, enabled by integration with Amazon SageMaker Neo, makes optimized machine learning models run up to twice as fast and consume less than a tenth of the memory footprint. These models can be run anywhere in the cloud or at the edge.

AWS has also built specific blueprints that combine ML and IoT data for common industrial use cases such as predictive maintenance and predictive quality.




IBM was one the first of the large software vendors to put a stake in the ground with regards to ML. In 2011, its Watson computer system won the quiz show ‘Jeopardy!’ It had been using ML in small ways since at least the 1980s.

With advanced AI tools, Big Blue takes a platform approach targeted at developers. Its set of APIs (spanning language, speech, vision and data) are meant for developers to use as the basis of AI-powered applications.

IBM also has Watson-branded tools; for instance, it offers Watson Analytics for cloud-based automated data analysis, visualization, and predictive and prescriptive analytics.

IBM brings its AI competencies to the IoT opportunity via Watson IoT Platform, its end-to-end cloud managed service for device management, connectivity, message, data visualization and storage. IBM also plays in the ‘comes with’ space with products such as Maximo Enterprise Asset Management and TRIRIGA (facilities lifecycle management), which includes both IoT data and AI/ML.




Google is working hard to prove that it’s ready for enterprise business. It has vowed to meet enterprises where they are and provide services that can be directly adopted.

Google is rolling out tools to simplify the use of ML and links to IoT. It has made impressive progress, which has gained it the right to be included in technology-planning decisions for many. Google Cloud Platform (GCP) has added the last few pieces to complete its IoT platform for enterprises with the unveiling of its Edge offering, putting it on par with rivals Microsoft and AWS.

The GCP IoT unit’s Cloud IoT Edge software and Edge TensorFlow Processing Unit (TPU), an edge-oriented processing unit for executing ML, bring AI to the edge while supporting GCP’s overall hybrid IT offering.

The GCP IoT approach nests within Google Cloud’s overall ‘AI first for business’ and hybridization themes. IoT is the next big thing for GCP, because it complements its analytics-first stance on delivering value from increasingly large datasets, along with Google corporate’s engineering-driven culture. GCP IoT will continue adding features that enhance its ability to crunch through data and generate outcomes.




Microsoft has established itself as the current #2 global cloud services platform, and has become a major player in meeting enterprise demand for advanced IT services and capabilities.

Microsoft has both the IoT cloud and edge covered with one of the most popular enterprise IoT platforms, Azure IoT.

To bring together its advanced technologies into IoT with AI, Microsoft has opened ‘IoT and AI Insider Labs,’ where it makes facilities, platforms and Microsoft developer experts available to customers and partners for co-innovation. Labs are located in Redmond, Washington; Shenzhen, China; and Munich, Germany.

Microsoft has also developed a set of prepackaged Azure IoT solution accelerators that help customers deploy and customize common IoT scenarios such as remote monitoring, predictive maintenance and device simulation.


IoT + OT Platforms




Bosch is both an end user and a vendor of industrial IoT given its breadth of OT market offerings ranging from automobile components to industrial automation.

Bosch is also a driving organization within the German Industry 4.0 initiative, the Industrial Internet Consortium and the IoT working group of the Eclipse Foundation.

The company’s IoT offering, IoT Suite, is a combination of managed services ranging from device connectivity, digital twin, and gateway software and management to analytics. The services can run in the Bosch IoT Cloud or in AWS with a subset of services for edge and remote management available on SAP and Microsoft Azure as well as on premises.

The Bosch Center for Artificial Intelligence was founded in early 2017 to deploy cutting-edge AI technologies across Bosch products and services. Bosch brings IoT and ML together via the Bosch IoT Analytics Track and Analyze offering, which was codeveloped with partner Software AG.




C3 IoT has positioned itself as the platform provider behind large enterprises’ digital transformations by combining the functions of a classic IoT platform, including device management and data capture, with business data integration using AI and ML.

The C3 IoT platform, which includes data-manipulation tools, provides the underlying functionality to tie together multiple IoT implementations across a given customer business.

The C3 Type system – the core of the platform – is the data-abstraction layer that models all elements within the entire system. These data models are accessed and manipulated by application developers with a RESTful interface or through role-specific C3 IoT UIs such as C3 Ex Machina to build and visualize analytics solutions.

The architecture is designed to support and interface with other systems such as Jupyter Notebook and leverage existing ML with TensorFlow and Spark MLlib. Application developers can use an Eclipse IDE plug-in, in addition to support tools such as a UI designer and a visual C3 Type designer.




Hitachi Vantara offers Lumada as a core IoT platform with higher-level functions such as data ingestion and blending, orchestration, data analytics and visualization.

Hitachi Vantara’s Pentaho is designed to address all aspects of the data engineering, data preparation and analytics pipeline – including data ingestion, processing, blending, delivery, discovery and analytics. ML orchestration capabilities are core features of Pentaho.

The heavy industry arena is one well known to Hitachi, specifically in mining and construction, as well as the utility sector (water and electricity). In Europe, the transportation sector, including rail systems, continues to be a major focus.




GE Digital, having faced slower-than-expected uptake of Predix, has refocused on its core constituency.

Following suit with GE corporate, GE Digital has become more focused on how it addresses its opportunities, choosing to focus on quality and depth of opportunity over quantity of engagements. To that end, the company will focus on vertical markets, including energy, oil and gas, healthcare, transportation and mining.

GE Digital launched its data analytics offering, Predix, in 2013. The company has expanded the purview of Predix with additional analytics packages focused in key GE verticals such as power and industrial systems, including launching its own cloud offering, Predix Cloud, in 2015. Built on Cloud Foundry, Predix is synonymous with application analytics platforms in the industry, and is the yardstick against which many other platforms are measured.

GE is also now supporting both AWS and Azure for infrastructure to run Predix, which is a pivot from its own failed attempt to build a GE Predix Cloud.

M&A by GE has included data-modeling company BitStew, ML firm and services firm ServiceMax. It is also a founding member of the Industrial Internet Consortium, alongside partners including Cisco and Intel. GE is an example of both a platform and ‘comes-with’ ML with Service Max, as it announced it can improve field service time estimates using the Apache Spark AI engine.




PTC is still relatively young in IoT, having bought its way to the table with $500m spent since 2013 on M&A, and is maturing on multiple fronts. The company’s overall business model transition from discrete software licensing to subscription-based sales has been well-executed.

To anticipate customer needs, PTC has pivoted its positioning from IoT platform provider, led by the ThingWorx suite, to a central creator of value for connected industries with a particular focus on manufacturing (specifically the factory).

PTC has native ML capabilities as part of the ThingWorx platform (ThingWorx Machine Learning). PTC will also support Microsoft in pairing its Azure IoT and Azure Cloud offerings with customer outcomes via Azure’s IoT Hub and PTC’s ThingWorx integrations.

As part of a strategic relationship with Rockwell Automation, PTC will integrate its ThingWorx, Kepware and Vuforia product suites with Rockwell’s FactoryTalk Manufacturing Execution Systems (MES), FactoryTalk Analytics for industrial IoT (IIoT) implementations, and Industrial Automation tools.




Schneider Electric, the 182-year-old French industrial firm, has its roots in industrial automation and the energy industry.

It is not surprising that when it came time for the company to launch its own IoT application platform (EcoStruxure) in 2017, it launched with analytics modules focused on its core competencies in building energy management, factory and equipment management, datacenters, and power generation and distribution (grid). The company is partnering with Microsoft and Accenture for AI enablement and building digital services that focus on IoT outcomes.




Siemens launched its MindSphere IoT platform offering in 2016. It was initially hosted exclusively in the cloud but has since expanded to include on-premises deployments.

The company has a multi-pronged approach to the IoT platform market with a packaged digital application store and DevOps support. In early 2018, the company released its third version of the Cumulocity-based MindSphere, which moves to include connectivity to almost any device or application, an essential capability for any IoT platform in the market.

The latest release of MindSphere is targeting the DevOps sector and includes broad alignment with AWS, giving access to native services with the platform. This is a strong signal of Siemens’ willingness to cross the OT chasm to align with IT best practices.

Siemens has announced plans to launch a new digital industries business unit in April 2019, which will focus on bringing edge computing and AI into the heart of the company by using MindSphere to support its diverse businesses including rail, process industries and digital factories.

Going forward, Siemens will consist of three operating companies – digital industries, smart infrastructure, and gas and power – as well as three strategic companies: Siemens Healthineers, Siemens Gamesa Renewable Energy and Siemens Alstom.

IT – Application-Centric




The OpenText IoT platform stack includes identity and security, IoT analytics, IoT applications and the Core IoT platform. In addition, the OpenText business network and services can be employed to add significant value for enterprises seeking to trade, integrate and communicate with B2B partners.

The OpenText platform enables the creation of digital twins, low-code IoT process and workflow development, and detailed access control capabilities. Guidance Software (acquired in 2017) brings forensic endpoint security capabilities (primarily used on laptops today) to IoT use cases. At the intersection of IoT and AI, the company positions Magellan.

OpenText launched Magellan ML platform to infuse machine-driven intelligence and analytics into its entire stack (including IoT), a move that is intended to underpin the company’s game plan to place advanced analytics into the hands of every user. It makes use of Apache Spark’s MLlib library of machine algorithms, and also draws on Spark and Hadoop for data processing.

Currently enterprises, are using the combination of OpenText IoT platform and Magellan for use cases such as predictive maintenance.




Oracle is attempting to make up ground on AWS and Azure by adding datacenter locations, features and performance to its Oracle Cloud Infrastructure, but it’s far behind.

In 2019 and 2020, the company intends to introduce nearly 20 new datacenters globally, including one per month in 2019.

In IoT, it will compete by translating the underlying features and performance into IoT applications and, ultimately, business outcomes. Oracle’s IoT group argues that, for a given company’s IoT implementation to be successful, the focus needs to be on business outcomes, not underlying technology decisions and IT integration.

Oracle’s technology to support IoT includes a framework and set of capabilities that form layers, starting with its cloud as the basis or platform layer, followed by its five core applications, and technology and industry specialization on top.

Oracle’s strategy for leveraging ML is centered on what it calls Adaptive Intelligent Apps (AI Apps). These are applications that are being enhanced with ML functionality to enable users to make better decisions based on insights derived from large datasets.

The company’s asset management approach serves as one example of its overall approach. With its Asset Monitoring app, it combines the ability to track and monitor asset performance and health with digital twins, the virtual representation of a physical asset, and digital threads, which tie predictive modeling inherent in digital twins into business processes.




SAP is taking a similar application-led approach to Oracle, with about 50 ‘intelligent’ applications of its own. However, it has also launched a set of ML business services directly embedded into SAP applications.

SAP also has Intelligent Enterprise offerings that take advanced technologies (such as IoT and ML) and couple them with SAP’s business application software to drive positive business outcomes for customers. It is a sensible approach given the vast number of SAP business users, as opposed to those who possess ML skills and are also working within SAP environments – although SAP is working hard to expand that cohort of developers. SAP offers parallel platform offerings for customers who do have their own ML experts and need to run ML in an SAP environment.

To address IoT, SAP has announced SAP Leonardo IoT, a multi-purpose platform for IoT enablement of business processes. For workloads requiring edge services including ML, SAP Leonardo IoT Edge can be deployed on-premises or in the cloud.

SAP offers several paths for customers based on needs: combining IoT data with business process data, extending existing SAP IoT applications, or building completely new solutions. At Mobile World Congress 2019, the company announced cloud-to-cloud interoperability between SAP Leonardo IoT and Microsoft Azure IoT Hub.




SAS, an ML and analytics leader, has built a platform for specifically addressing the intersection of AI and IoT called SAS Analytics for IoT. The solution offers support for centralized or edge execution venues and integrates streaming analytics, real-time data management and visualization.

The latest release of SAS Analytics for IoT (7.1) also supports container deployment and a streamlined approach to analytics model management. The company has significant incumbency in IoT-friendly vertical markets including transportation, healthcare, retail and manufacturing.

The company has chosen a ‘works with’ strategy for cooperating with public cloud providers including AWS and Azure.




Software AG sits at the heart of the Industry 4.0 movement in Germany and continues to build on the IoT platform assets it acquired when it purchased Cumulocity in 2017.

That platform was focused on device connectivity and management across a range of network protocols, integrated data analytics of the devices, and APIs for integration with third-party visualization tools and BI systems. It can provide a single point of authentication for a heterogeneous device environment. The platform delivers native support for a wide variety of popular industrial communication protocols including Modbus, OPC UA and CAN bus. Software AG works with Zementis to power the Software AG Artificial Intelligence Solution, which can be combined with its IoT platform.