IoT energy efficiency: Machine learning from long-term habits

IoT energy efficiency: Machine learning from long-term habits
Rising energy costs is a reality: utility bills are increasing, and this can be a concern for a lot of homeowners. Using (or even wasting) more energy than needed is problematic for the environment, but also for our wallets.

Smart home energy tips can come in multiple ways. It includes automated schedulers that shut down electronics overnight, or sensors detecting motion allowing lights to turn on when people enter a room and off when people exit. This type of tips is becoming quite common in households, but can definitely go one step further.

A smart power system in a home would also monitor and manage how and when power is consumed. It could be used to control the amount of time kids spend on their electronic devices and to turn off power-consuming appliances or systems unless they are in use. A green, smart house goes beyond turning the HVAC system on and off as scheduled. It automatically opens and closes window shades or curtains to adjust for the sun and the season.

The smart home’s meter charges the home’s power storage system during the day via solar panels on the roof, or at night when power is less expensive. That way, the home’s power-greedy appliances can use “cheap” stored electricity instead of drawing from the grid during expensive rate times. These systems are already in use in industrial applications and will soon be moving to home use.

An effective smart home would also make its inhabitants smarter. When people are educated about how much appliances are costing in power, and how they can change their habits, they are more likely to turn off the appliances when not in use and to be more energy-thrifty when they do use them. For instance, it’s much more efficient to run washing machines or dishwashers when filled to capacity. Running appliances under capacity wastes power and water, not to mention detergents.

Cost-effective comfort and convenience

By using a network of position and motion sensors, the smart home controls temperature and lighting to accommodate which rooms in the home are being used at any given time.

For instance, in a smart lighting use case, if someone comes home with the car at a regular hour every day, the garage could be already lightened up when he arrives, and even the paved way too, if it is dark outside. If the family is watching a movie, the system would be smart enough to turn off the lights in the empty parts of the home, especially in areas that are usually not visited during those hours. When the family goes to bed, the system can then turn on lights in bedrooms and bathrooms and dim them progressively in the kids’ bedrooms to prepare bedtime, and even leave it on a bit more in the hallway for the last one who is afraid of the dark.

The analytics behind the smart system are key. Older systems with thermostats or using simple IFTTT technology are difficult to program and too rigid, to make exceptions. The newer systems learn from the people who live in the home to make predictions about future behaviors — the number of household members, how rooms are used and when, bedtimes, who works from home and where, who gets up early, etc. Of course, simple overrides are available when needed, but the majority of patterns are absorbed by the system and used to enhance comfort and convenience settings that also happen to be cost-saving.

Beyond utility bills – Predict the irrevocable

There are also some less obvious energy efficiency advantages that the IoT can offer to businesses and consumers. What about unexpected expenses like accidents and equipment failures? Let’s look at water heaters, for example.

When a water heater starts to go bad, it’s always in the form of a slow leak. This type of equipment failure can be tricky to identify. If not immediately detected, the cost of the leak can quickly add up. In many cases, the water heater continues to run, inefficiently heating the water that is traveling to the dishwasher or shower, along with the water that is leaking out. This runs up the utility bill without the homeowner receiving the benefit of enjoying all the hot water. The simplest fix is to install a water heater leak detector that sends an alarm when the tank fails. But by taking that one step further and connecting that leak detector sensor to a smart home network that includes actuators and controls on the power and water sources, the smart home can alert the homeowner and control the power and water systems.

Of course, water heaters can also fail in the form of a tank rupture that spills gallons of hot water, floods the house and creates costly damage. The same type of damage can result from a frozen water pipe that breaks. A smart home with a water flow sensor can be programmed to notice when water is moving in the pipes when no one is home. It can send a notice to the homeowner and turn off the water at the main valve, saving valuable resources and avoiding high water bills and expensive flood damage.
 

When energy efficiency extends to financial efficiency

Most of us recognize the advantages of the smart building or home when it comes to security and convenience, and even increased energy efficiency. But in addition to the financial benefit of cutting utility expenses, truly smart energy management reduces the possibility of damage from natural and unnatural disasters, reduces insurance costs, educates management and homeowners about power usage, and increases comfort and convenience while saving money.
 
Cees Links, General Manager,
Qorvo Wireless Connectivity
Business Unit

*Contributed by Cees Links, General Manager, Qorvo Wireless Connectivity Business Unit

Cees Links is a pioneer of the wireless data industry, the father of Wi-Fi and the man who led the team that created and popularized Wi-Fi. Links was the founder and CEO of GreenPeak Technologies, a smart home and IoT radio communications semi-conductor company, now part of Qorvo. Links was recognized as Wi-Fi pioneer with the Golden Mousetrap Lifetime Achievement award.

Share to:
Comments ( 0 )