What is a Smart Thermostat? Algorithms

There is no clear definition of a smart thermostat. In fact, as more offerings have entered the market, the definition of a smart thermostat has become more misunderstood.

Some consider a thermostat “smart” if it has two-way communications (i.e., you can view and change settings on your smartphone). However, this is simply inaccurate. This type of thermostat, according to Navigant Research, is actually considered a communicating thermostat. A communicating thermostat is a device solely enabled to gather and transmit in-home temperature data in a two-way format that can be accessed and adjusted remotely via a web portal or mobile application. Unfortunately, a handful of communicating thermostats on the market today have actually been labeled or marketed as smart thermostats despite their lack of smart features, and the terms communicating, smart, and even programmable have become increasingly interchangeable.

So what is the difference? And what is it that makes a smart thermostat smart?

What makes a smart thermostat smart is algorithms. Algorithms hidden in the back-end software of a device are what enable the device to perform a variety of different functions that make it advanced, and not just communicating. For thermostats specifically, algorithms can optimize HVAC settings for consumers.

There are algorithms for a variety of different thermostat functionalities. For example, a Nest thermostat’s learning algorithm is what enables the device to adapt to a user’s lifestyle and the changing seasons to optimize HVAC settings and minimize energy consumption without affecting comfort. EcoFactor uses a series of advanced algorithms alongside real-time data from communicating thermostats, weather conditions, and customer preferences to make automatic micro-temperature adjustments to save customers money and energy. The ecobee 3 is advertised as smarter due to its support of algorithms for capabilities ranging from alerts and reminders to HVAC system recovery to optimal humidity control to home/away sensing.

Several common functionalities of thermostats enabled by algorithms include:

• Geofencing: A feature that uses the global positioning system (GPS) or radio frequency identification (RFID) to define geographic boundaries and create virtual barriers where triggers can be set up to send alerts when a device enters or exits (i.e., when a user enters the vicinity of his or her home, the thermostat receives an alert and activates the HVAC system).
• Upstaging and downstaging: A technique for multi-stage HVAC equipment in which stages of heating or cooling are automatically initiated based on the difference between desired temperature and actual temperature. A higher stage will automatically engage if actual temperature is significantly different from desired temperature, and switch to a lower stage when the actual temperature is closer to the desired temperature.
• Free cooling: For users that have an economizer or ventilator, a feature that can detect when outside temperatures are cooler than the air inside a consumer’s home and use the outside air to cool the home instead of running an air conditioning unit.

Algorithms are still improving and increasingly support more advanced capabilities. Manufacturers of thermostats have teams dedicated to developing new algorithms and adjusting existing ones to best serve customers. And similar to updates in apps like Instagram, these algorithms can be added or modified on already installed devices. What developers can do with algorithms and the value they add make these devices unique and desirable to consumers. Without Nest’s algorithms, for instance, the device would only be a well-designed on-off switch, controllable through a smart phone, not the smart, learning thermostat that has driven the market and become one of the most popular offerings available.

Differentiating between smart and communicating

Despite how straight-forward a smart thermostat may seem, the difference between smart thermostats and communicating thermostats is still commonly misunderstood. This is in part due to the shared features between smart and communicating thermostats. Indeed, smart thermostats actually are communicating thermostats—the only difference is added algorithms. This means that both communicating and smart thermostats have capabilities for demand response (DR), enhanced programming and scheduling, data availability (both can store and transmit data), and remote control capabilities.

Additionally, both smart and communicating thermostats may have accompanying web portals and mobile apps that provide insight into a user’s energy consumption and tips for increasing energy efficiency and reducing monthly bills. Misunderstanding also comes from the fact that no regulatory body or authority has defined “smart.” The Environmental Protection Agency has not created a definition for smart thermostats, and there is no regulation over the use of the term. This can have negative implications on the market, from frustrated consumers to inaccurate quantification of these devices.

For simplicity’s sake, Navigant Research categorizes smart thermostats as those with the same functionality as a communicating thermostat (i.e., two-way communication), but enhanced by data gathering and analytics that optimize HVAC settings for efficient and automated energy consumption. Where greater specificity is warranted, Navigant Consulting has defined a smart thermostat as having at least three of the following features (enabled by algorithms): occupancy detection, heat pump lockout temperature control, upstaging and downstaging optimization, optimal humidity control/AC overcooling, fan dissipation, behavioral features, and free cooling/economizer capabilities.

To learn more about smart thermostats and how they are different from other thermostats, look for Navigant Research’s 2Q 2016 Smart Thermostats report.

Paige Leuschner is a research analyst with Navigant Research in the Utility Transformations program, contributing to the Residential Energy Innovations research service and supporting the Demand-Side Management research service. Paige specializes in advanced metering infrastructure, smart thermostats, home energy management, and utility customer engagement.