IoT + AI, or how to reduce energy costs

In an era of rising energy prices and increasing environmental awareness, companies are looking for effective ways to optimize their energy consumption. The combination of the Internet of Things (IoT) and Artificial Intelligence (AI) is opening up new opportunities for companies looking to reduce their bills and contribute to sustainability. How can IoT and AI technologies contribute to more efficient energy management? In this article, we discuss specific benefits, challenges, and real-world IoT use cases.

IoT - table of contents

What is the Internet of Things? (IoT)

The Internet of Things, also known as IoT, is a technology that enables the connection and exchange of data between various devices connected to the Internet. It is a network of physical objects – “things” – equipped with sensors, software, and technologies that allow them to exchange data with other devices and systems over the internet. But why connect all devices to the Internet – from printers and refrigerators to air conditioners, trucks, and industrial machines?

Research conducted by Transforma Insights indicates that the production of new IoT devices will increase global electricity consumption by 34 terawatt-hours (TWh) by 2030. Meanwhile, training and using large language models, such as ChatGPT, may lead data centers to draw up to 21% of the world’s electricity demand.

However, Internet of Things solutions will also reduce electricity consumption compared to using the same type of sensor-less devices by over 1.6 petawatt-hours (PWh). To illustrate the scale – that’s the amount of electricity needed to cover the consumption of over 136.5 million homes for one year. Businesses, however, cannot thrive without increasingly new devices, and artificial intelligence can greatly assist in further minimizing the energy they consume.

Source: Transforma Insights (https://www.6gworld.com/)

What is the connection between IoT and AI?

Artificial Intelligence analyzes data from Internet of Things devices, identifying trends and strategies to save energy. Thanks to this, IoT and AI open up new possibilities for efficiently managing energy consumption. For example, AI can adjust Heating, Ventilation, and Air Conditioning (HVAC) systems based on:

patterns of room usage – for instance, lowering the temperature in an unoccupied room on a given day,
weather conditions – reducing the air conditioner power on a cloudy or windy day,
indoor air quality – increasing ventilation efficiency only during a meeting involving many people.

Additionally, AI supported by IoT enables the creation of digital twins to optimize manufacturing processes, such as predicting machine part consumption or analyzing the frequency and causes of failures leading to downtime.

IoT AI technologies are beneficial for the entire electricity supply chain – consumers, distributors, and power plants. Internet of Things allows the automation of various processes, such as collecting energy data for measurements and analysis. The application of hardware power constraints and the optimization of AI model training time can reduce energy consumption by approximately 12% to 15% with minimal impact on model performance.

How can IoT and AI reduce energy costs in your company?

IoT sensors in buildings send real-time energy consumption data for analysis, speeding up repairs and reducing energy losses due to faulty equipment. Here are examples of sensors that can be installed in buildings:

temperature sensors – monitor temperature changes in rooms and allow for optimizing energy consumption for heating or cooling,
light sensors – detect the level of natural light and automatically adjust the intensity of artificial lighting,
motion sensors – enable system activation only when people are present in the room, reducing unnecessary energy consumption,
smart meters – collect data on energy consumption, enabling detailed analysis and optimization,
air quality sensors – regulate the operation of ventilation systems, ensuring proper air quality with minimal energy consumption.

Smart thermostats and lighting systems can adapt to room usage patterns, reducing energy costs. IoT devices can decrease electricity consumption and costs in buildings during peak hours through demand response programs and scheduling. Implementing IoT and AI-based solutions can lead to energy savings of up to 40%. Furthermore, AI-controlled energy storage solutions optimize its use by managing storage and distribution.

Source: DALL·E 3, prompt: Marta M. Kania (https://www.linkedin.com/in/martamatyldakania/)

Benefits of using IoT and AI

Introducing intelligent IoT and AI technologies brings many benefits. It’s not just a way to lower bills but also to enhance the comfort of office use, primarily:

better temperature control and energy savings thanks to smart thermostats,
control of lighting brightness, which means a significant reduction in electricity consumption, especially during the autumn-winter period,
air humidity management – particularly important when storing documents or sensitive, specialized equipment.

IoT devices are also becoming increasingly sophisticated, and in response to growing demand, sensor prices are decreasing. An example is Flex2X, a solution created by the British company Grid Edge (https://gridedge.ai/). This system integrates information from already installed sensors in buildings, such as temperature or humidity detectors, with additional data, such as weather conditions. It then analyzes them using AI algorithms, continuously optimizing energy utilization in the building.

Source: DALL·E 3, prompt: Marta M. Kania (https://www.linkedin.com/in/martamatyldakania/)

Downsides of IoT – how much energy does AI require?

Unfortunately, the combination of Internet of Things and AI comes with its drawbacks. The use of IoT and AI is primarily associated with the need for significant energy consumption. Detailed information on the amount of energy required to train the most powerful and well-known GPT-4 model, which is used in the commercial version of ChatGPT and BingChat, has not been publicly disclosed. However, based on available data, we can outline the scale of this undertaking.

The GPT-4 model has over 175 billion parameters and was trained on a dataset exceeding 45 TB. This requires an intensive process of data analysis and model parameter tuning, leading to a significant demand for computational power and resulting in high energy consumption.

To train the GPT-4 model, advanced Graphics Processing Units (GPUs) and Tensor Processing Units (TPUs) were used, known for their intensive energy consumption.

Summary

IoT and AI can have a significant impact not only on a company’s financial results but also on the environment by improving energy efficiency. The average large building consumes approximately 230,000 kWh of electricity annually, translating to a cost of around $25,000. The implementation of Internet of Things can contribute to reducing energy consumption by 10% to 15%. The integration of IoT and AI is also crucial for fully harnessing the potential of smart factories in terms of sustainable development and cost savings.

If you like our content, join our busy bees community on Facebook, Twitter, LinkedIn, Instagram, YouTube, Pinterest, TikTok.

Author: Robert Whitney
JavaScript expert and instructor who coaches IT departments. His main goal is to up-level team productivity by teaching others how to effectively cooperate while coding.
View all posts