More Topics
Paid Content : This paid content was written and produced by RV Studios of Red Ventures' marketing unit in collaboration with the sponsor and is not part of ZDNET's Editorial Content.

Machine learning and the Internet of Things

Gartner predicts that more than 65 percent of enterprises will adopt IoT products by the year 2020.

The growth of the Internet of Things (IoT) market in recent years is hard to ignore. According to Forbes, the global IoT market will grow from $157 billion to $457 billion between the year 2016 and 2020. The major contributors to the investment include leading industries like manufacturing, logistics, and transportation.

When it comes to sectors that dominate this investment, smart city initiatives and industrial IoT top the chart by owning more than 50 percent of the market. Gartner predicts that more than 65 percent of enterprises will adopt IoT products by the year 2020.

A typical IoT solution pipeline consists of the following five stages:


The heart of this process and what drives the real business value is encapsulated in the third stage of this activity chain, which is 'Transformation and Analytics'. This is the stage where the data is inspected and decisions are made. These decisions will directly influence the actions that will optimise business flows.

This is where the role of machine learning and artificial intelligence becomes significant. The ability of the system to make cognitive decisions based on historical data will greatly influence the value of the solution. Technologies like Azure Machine learning can leverage supervised learning techniques to help make business decisions based on classification, regression, and anomaly detection.

Machine learning - evolution

The concept of machine learning is not new to the world of computing. The birth of the term happened in the late 1950s, inspired from related fields in computing such as pattern recognition and artificial intelligence. However, leveraging this concept to optimise business process was largely constrained by the cost of provisioning and maintaining the compute and storage required to host and execute machine learning algorithms.

The primary cause for the re-emergence of machine learning is the evolution of cloud computing and its adoption in today's enterprise world. By offering features like infinitely scalable compute and storage, high-performance computing services, and pay-per-use subscription model, cloud computing became the ideal surrogate to bring machine learning back to life. This enabled organisations of any scale to affordably run machine learning algorithms to optimise their business processes. It also encouraged cloud market giants like Microsoft, Amazon, and Google to offer this technology as a software service consumable on a subscription model.

Machine Learning and IoT

Machine learning uses supervised learning techniques on historical data to make cognitive decisions. The greater the quantity of historic data, the better the decision-making capabilities of the algorithm. This philosophy makes IoT the ideal use case for machine learning as the data generated by the devices are usually very frequent.

The following are few common scenarios where machine learning works hand-in-hand with IoT to enable business optimisations:

  • Anomaly monitoring -- Azure machine learning can be used to detect anomalies in time series data, in data feeds sent by the IoT devices that are uniformly spaced in time. Anomalies like spikes and dips, positive and negative trends, can be detected using a machine learning algorithm monitoring the live stream of device feeds.
  • Predictive maintenance -- Predictive maintenance directly impacts the costs for an organisation, which makes it one of the most popular machine learning solutions. The ability of machine learning algorithms to foresee possibilities of a device failing, remaining life of an equipment, and causes of failure can enable the business to optimise operational cost by reducing the maintenance time significantly.
  • Vehicle telemetry -- The capability of machine learning solutions to ingest millions of events from vehicles to improve their safety, reliability, and driving experience makes it a desirable technology to adopt for transportation and logistics industries.

Microsoft technology stack for Machine Learning and IoT

Amongst the popular Cloud providers, Microsoft was the first to launch a full-fledged IoT and machine learning solution. The offering from Microsoft involves multiple technologies offered as a service to cater to different phases of the IoT pipeline. Following are few of these technologies:

  • AzureIoTHub
  • AzureIoTsuit
  • AzureIoTEdge
  • AzureEventHub
  • Azure Stream analytics
  • Azure Machine learning
  • Microsoft cognitive services
  • Supporting technologies
    • Azure Event Hub
    • Azure Service bus topics
    • Azure event grid

More about these technologies and the role they play in the IoT pipeline will be covered in my next blog.

Read more about IoT for the enterprise.

Editorial standards