The Foundations

Within the next 12 months, IoT Analytics predicts that for every human being on earth, up to six devices (excluding smartphones or computers) will be connected to the Internet. Such interconnection of devices is termed the Internet of Things or IoT. This network of physical objects contains a wide array of embedded technologies, allowing them to communicate, sense and interact with their internal or external environment. These tasks are often performed without human interaction.

IoT devices come in many shapes and sizes, from mobile phones, smart speakers, lightbulbs and fitness trackers to sensors that can detect light, moisture, movement, temperature, tilt, and pressure. For example, in the City of Melbourne, a series of people-counting sensors are installed at key locations around the city. This network of sensors enables the tracking of pedestrian movement at macro scale. This allows the city to determine the vitality of an area at different times of the day or even days of the week. This data is useful for urban planning by guiding decisions relating to walkability and major events management. This in turn aids the city when providing guidance to local businesses, helping them manage retail staffing and security.

One particularly valuable aspect of IoT lies in the significant volume of data generated. IDC predicts that IoT devices worldwide will generate 79.4 zettabytes (ie. 79.4 billion terabytes) of data by 2025. That is the equivalent to the data generated by the Hubble Telescope if it operated for 7.94 billion years!

With this much data, a higher level of analysis is required to generate value. Big data analytics makes it possible to get timely answers from this sea of data. It helps uncover hidden patterns and generates insights from an unfathomable quantity of data. An example of big data analytics in action is the music recommendations to individual users by Spotify. The system uses the data generated by its 271 million monthly active users to suggest personalised songs of interest.

Another example is its application in the finance and securities industry where big data analytics is used to identify fraud, money laundering and illegal trading. The sheer volume and variety of data could not have been processed efficiently or as rapidly through traditional means.

Applying machine learning to analysis generates greater efficiency. Machine learning (ML) enables machines to learn by themselves using the provided data. The premise is that a system can learn from the data it is analysing and identify patterns. As more data becomes available, machine learning algorithms improve their results and adapt to gradual changes automatically. A prominent consumer application of machine learning is the smart assistants used in phones and speakers like Siri (Apple), Alexa (Amazon) and the Google Assistant.

Big data means big analysis tasks. Big analysis tasks are more efficient with automation. The combination of machine learning and big data is enhanced with Artificial Intelligence (AI). AI helps make decisions with minimal human intervention. This result is software that can learn from experience, adjust to new inputs, and perform human-like decision-based tasks.

Through a combination of big data analytics, machine learning and AI, machines can be trained to complete many more tasks. These tasks would have once required the involvement of a human to make a decision.

These machines can interact with other devices, generate data, process data and recognise patterns automatically. The total effect is independent decision-making by machines.

Rising Demand

5.8 billion

IoT endpoints by 2020

(Garter, 2019)

1 trillion USD

Worldwide spending on IoT by 2022

(IDC, 2019)

28 billion AUD

IoT spending in Australia by 2021

(PwC, 2018)

Demand for IoT alone is expected to boom. This interest has been driven by a number of factors:

Monitoring and Control

IoT networks enable the comprehensive monitoring of the condition, operation and usage of products and equipment. Monitoring and identification of changes in real-time enables autonomous alerts and notification. For example, smart devices can be used to:

  • Monitor machines for predictive maintenance in manufacturing environments
  • Track the location of inventory in retail, logistics and transportation industries
  • Observe patient health and generate alerts in the healthcare sector

Optimisation

Monitoring and control enables algorithms to optimise how products, equipment or systems operate. Examples include:

  • Optimising passenger flow and route efficiency in transportation
  • Managing city waste collection in real-time using bin level sensor data to optimise truck routes so they only visit bins that need emptying

Autonomy

Combining monitoring, control and optimisation with AI enables autonomous systems to complete machine-driven decision-making. Examples of this are:

  • Managing self-driving vehicles to transport precious minerals in the mining industry
  • Irrigation systems paired with moisture sensors and crop condition monitoring to determine the optimal levels of watering for specific topographies, seasons and times of day in the agricultural sector
  • Autonomous product operation, enhancement and personalisation

Creating Competitive Advantage

IoT solutions complemented by big data analytics, machine learning and AI technologies can open up new business opportunities, identify new revenue streams and create new business models.

IoT is an incredible enabler. But without the right platform and partner, an IoT solution can become more complex than the problem it is meant to solve.

Before starting on the path to IoT, ask the right questions first. Are you clear about the problems you have? What improvements you need? Do you need any assistance in getting started on an IoT and data analytics project?

With a partner that has the right IoT skills, experience, and capability, your journey towards digital transformation is more likely to succeed.

Dobrica Vasic
Manager - Research and Development
dobrica.vasic@nec.com.au

Additional Contributors:

Rajesh Chopra
Test Lead
rajesh.chopra@nec.com.au

Thirukkumaran Sivahumaran
Senior Software Development Engineer
thirukkumaran.sivahumaran@nec.com.au

Peter Watson
Subject Matter Expert
peter.watson@nec.com.au