Get To Know The Four Types Of Data In The Internet Of Things

Guest author David Friedman is the cofounder and CEO of Ayla Networks.

Big data is one of the greatest economic opportunities of our time.

It’s also incredibly vague. You’ve probably been in conversations where the different participants were using “big data” to refer to (a) large amounts of data; (b) data sets that exceed the capabilities of traditional databases; or (c) the software tools employed to analyze the data sets in the first two definitions.

One of the most significant benefits of the Internet of Things is the fact that it will vastly expand our ability to monitor and measure things taking place the real world. A shop-floor manager knows that a small whining sound in a motor might mean trouble. A typical homeowner will also know that the ventilation system in your dryer can clog up with lint and create a safety hazard. Data systems are finally giving us an ability to understand these problems with precision.

The challenge, however, is developing systems and business models that make that information valuable. Think of smart thermostats for a moment. During a peak power emergency, utilities and third party energy service companies will want exact updates on power consumption every minute: Being able to precisely fine-tune power and maximize savings can be the difference between a normal summer day and a debilitating brownout. But between the hours of midnight and 4 am, the need for information is far less urgent: The data will mostly be valuable to determine long-term trends.

Now think from the consumer’s perspective. Data updates even at, say, 15-minute intervals would lead to overload. It wouldn’t just be worthless: it would create nuisance that would detract from its value. Instead, consumers probably just want a monthly summary which points out a few trends.

I talk to people all the time about the “data value” challenge. The below list is a summary of the general categories of data and the opportunities manufacturers and service providers are pursuing. 

The Five Kinds Of Big Data

Status Data

Are the air compressors for the cold storage unit working? Did one just suddenly drop in performance? Status data essentially providers consumers and or businesses with an ongoing EKG of the world’s things. 

Status data is the most prevalent, and most basic, type of IoT data. Virtually everything will generate data like this as a baseline. In many markets, status data will be mostly used as raw material for more complex analyses, but in many markets it will have a significant value of its own. 

Look at what Streetline had done for parking spots. The company has created a system that notifies subscribers about open parking spaces. Sure, the long-term data helps city planners, but to most consumers the immediate status data is the most important thing. 

Location Data

Where is my product? Did it make it to its destination? Location is a logical extension of GPS. GPS is great, but it doesn’t work well indoors, in crowded spaces or in rapidly changing environments. Someone trying to track pallets and robotic forklifts will want real-time information. 

Agriculture, which could become an early IoT market, will make extensive use of location data because owners have to track equipment across huge geographical areas. While we’ve already seen the debut of consumer products so people can locate their keys, a larger markets exists for serving commercial and industrial customers, particularly where there are numerous assets to track, few employees, and need to track things in real, and near-real, time. Developing the Foursquare for paint warehouses is a huge opportunity. 

Automation Data

Consumers are rightly skeptical about automation. You don’t want to be stuck in a dimly lit office or a chilly hotel room because some building management system care more about saving a few dollars than your comfort. Automation also creates security issues. 

Nonetheless, automation is inevitable. No one is going to sit with their finger on the thermostat to save $4.75. Likewise, lighting systems that depend on human interaction fail. (Some smart-lighting manufacturers want to use their sensor data to tell store managers when new checkout lines need to be opened.) The challenge will revolve around carving out applications and rules of conduct. 

Actionable Data

Think of this as status data with a follow-up plan. Buildings use 73% of the electricity in the country and a good portion—up to 30%, according to the EPA—is wasted. Why? Energy is a secondary issue for most building owners. They want to fix it but worry that the cost, time and headaches will outweigh the benefits. 

There are two ways around this problem: automation (see above) that can change the immediate state of a system, and persuasion that can get people to change their behavior or make long term investments. Opower has helped pioneer a solution to the persuasion problem by showing consumers and businesses how they compare to their neighbors along with data. According to their own studies, persuasion data can cut energy consumption by 2 to 3 percent. 

Creating A Feedback Loop With IoT Data

Do you know what your customers think? You may believe you do, but you’re probably wrong. In the near future, manufacturers will analyze data from their connected products to better understand how their products are used in the wild. Most companies right now have no idea how their products are used. They get shipped through distribution, bought at retailers, and end up at homes or offices. The user and the manufacturer rarely, if ever, communicate. 

IoT will create a feedback loop from consumer to manufacturer, where product builders will examine real-world behaviors—with the appropriate levels of privacy, security, and anonymity—to encourage continued product improvements and innovations.

Update: Due to an editing error, the “Automation Data” section was mislabeled.

Photo by Jakob Montrasio

Facebook Comments