In late December 2018, International Data Corporation (IDC) predicted that the collective sum of the world’s data would grow from the current 33 zettabytes to 175 zettabytes by 2025, much of it produced by the interconnected devices known as the Internet of Things (IoT). Unfamiliar with a zettabyte? To put it in perspective, IDC describes it like this:
“If one were able to store 175 ZB onto BluRay discs, then you’d have a stack of discs that can get you to the moon 23 times.”
I think we can all agree that’s a lot of data.
We can probably also agree that data is useless until it’s subjected to analysis. Until recently, organizations that wanted to perform data analysis had to spend time and money to develop their own data-crunching algorithms. Therefore it was a struggle for most companies to manage and analyze—and, as a result, gain benefit from—IoT-generated data.
Today, there are numerous pre-built IoT analytics solutions on the market, putting data analysis within reach of just about everyone. The availability of lower-cost IoT analytics solutions means more and more companies are able to present a stronger business case for trying out new IoT analytics applications.
[bctt tweet=”The availability of lower-cost IoT analytics solutions means more and more companies are able to present a stronger business case for trying out new IoT analytics applications.” username=”iotacomm”]
Below are some of the more interesting IoT analytics use cases we’ve heard about; if you know of others, tweet us @iotacomm. Keep in mind too, that, while most of the below examples have a specific industry slant, many companies—no matter what they do—are also starting to apply IoT analytics to their building operations for the purpose of reducing their energy bills. (You can read more about that here.)
6 Interesting IoT Analytics Applications
1. Social Analytics
According to Boris Shiklo at ScienceSoft, there’s been increased interest in social analytics, where companies leverage the IoT to gain actionable insights about an audience’s behavior and emotional response at sports matches, fashion shows, exhibitions, and other events.
During a “connected event,” sensors measure temperature and heart rate from a particular distance; video cameras monitor motion; and microphones detect cheering and booing. Machine learning algorithms applied to the data reveal the level of audience engagement and identify their emotional responses. The insights obtained with the analysis can be used for increasing future engagement, improving the effectiveness of marketing campaigns, and boosting brand awareness in social media.
Aquafarming (or aquaculture), the cultivation of plants and animals in water, is another intriguing IoT analytics use case. Nikolai Tenev, founder of DigidWorks, describes how the IoT is helping aquafarming operations gain more insight into their harvesting conditions. Using a variety of sensors, they can gather detailed data about the quality of the environment, including the temperature, nutrients, oxygen levels, light levels, and more. Analysis of the data can then help determine what is—and isn’t—good for the cultures, which helps them improve the water quality and ultimately boost production.
3. Predictive Maintenance
With the right data and the right IoT analytics tools, you can predict the future. That’s no small thing for businesses, who lose approximately $100,000 with every hour of downtime. No matter what industry you’re in, there’s an opportunity to improve operations using IoT-based predictive maintenance. Restaurants, for example, are using sensors to monitor their refrigeration units, allowing them to address potential problems before a failure occurs and food spoils. Manufacturers are doing remote equipment monitoring to avoid a breakdown of critical equipment that could impact the entire operation. Even office buildings are getting in on the predictive maintenance trend. In an effort to reduce the 190 million hours elevators are out of service each year, they’re putting IoT devices on them. Data analysis could, for instance, reveal if a door is opening more frequently than normal, which may indicate a problem—and a future trapped tenant. Wouldn’t you want to avoid that if you could?
4. Smart Roads
There’s a lot the IoT can do for travelers (like reducing traffic congestion and finding available parking), but these days, cities aren’t the only settings where the IoT can help. In mid-2018, the city of Denver, Colorado, began preparing to install concrete slabs embedded with sensors into a section of roadway as part of a trial run to see how certain data might be used on the state’s mountainous highways. The sensors will purportedly deduce things like the speed, weight, and direction of vehicles; that data can then be used to alert authorities to accidents. State officials have noted that certain areas of roadways, like a particularly dangerous curve on Highway 285, have high accident rates. Data analytics and the IoT make it possible to alert emergency responders the moment an accident occurs.
Aaina Bajaj of Signity Solutions points out a similar IoT analytics application: the ability to detect road maintenance issues. The ePave project, run jointly between researchers at the University of Buffalo and Chang’an University in China, is studying the idea of embedding sensors underneath road surfaces to detect moisture and pressure, which can indicate that problems are beginning to form. The idea is to detect road faults before they happen, and help motorists avoid traffic accidents due to damaged roads.
5. Smart Hospital Beds
Healthcare facilities are tapping into the IoT at an astonishing rate—one report states that 87% of healthcare organizations will have adopted IoT technology as of this year. IoT analytics applications in healthcare are many, everything from X-ray machines that use artificial intelligence to detect problems, to using asset tracking to stay on top of critical (and costly) equipment. Now, a new generation of hospital beds is using the IoT to provide higher-quality care for patients. The new beds continuously monitor a patient’s heart rate and respiratory rate; machine learning as part of IoT data analytics can then help detect signs of deterioration sooner. By some estimates, the technology inside the beds have been shown to help lower “code blue”-related mortality by 83% and cardiac arrests by 86%.
What is the AIoT?
The “Artificial Intelligence of Things” (AIoT) is when artificial intelligence technologies are applied to the data analysis process (and sometimes operational aspects of the network itself). The goal is to rapidly extract advanced insights from IoT data without human involvement. This is commonly done via machine learning.
Machine learning is a way of “teaching” a computer system to learn to perform certain tasks, such as making predictions or recognizing patterns in data. When applied to IoT data analytics, the system is capable of providing actionable information and improving decision-making.
Some IoT analytics solutions incorporate machine learning into their platforms; without it you’re less likely to experience the full benefit of the IoT. Still, there are human responsibilities that go along with advanced data analysis.
For instance, your organization will need to contextualize the data and define the goals and objectives of analysis: What business questions do you want to answer? You’ll also need to evaluate your company’s decision-making process to ensure you can keep up with the incoming flow of data. Real-time data is useful in the now, but it could become obsolete with the passing of too much time. (You can read more about the role of people in data analysis here.)
6. Precision Farming
Corporations aren’t the only ones looking to save money and increase efficiency; farmers are, too. In fact, the smart agriculture market is expected to grow rapidly through at least 2023. Precision farming is the practice of using IoT technologies to lower the cost of operation and improve farm performance. It enables farmers to deliver exactly the right treatment to plants and livestock, even down to individual animals and plants per square meter. For example, they can measure variations within a field of crops and make decisions about how to best apply things like pesticides and fertilizers, rather than having to apply them uniformly across the field. They can also monitor the location and health of their livestock remotely, which helps with the early identification of sick animals.