Starting with the thermostat
Of all household devices, the thermostat has one of the longest histories as an automatic control. Electric room thermostats were being patented in the United States in the 1880s, and by the early twentieth century they were used in many buildings to switch furnaces and boilers on and off. A thermostat is a very simple kind of automation: it measures a temperature and reacts when that crosses a set point.
The next step was to make the set point change by itself. A clock-driven thermostat could lower the temperature at night and raise it before morning, which saves fuel without anyone touching a dial. Mechanical timer thermostats of this kind appeared in the mid-twentieth century. The idea gained attention in the 1970s, when higher energy prices made saving fuel a public concern. Electronic versions, with a small display and memory for a weekly schedule, became common in homes from the 1980s and 1990s. These are what most people mean by programmable thermostats.
They did not always work as hoped. Setting a schedule involved a row of small buttons and a manual, and many owners overrode the schedule or never set it. This is a pattern in the history of home controls: a technology can be available long before it is easy to use.
Remote control by wiring and radio
While heating controls were becoming programmable, other household controls were getting a remote element. The power-line system X10, from the late 1970s, allowed lamps and appliances to be switched from a handheld or wall controller without new wiring. Infrared handsets for televisions spread in the 1980s, and some more ambitious systems tied lighting, security and audio together in custom houses. The early years of these ideas are told in the history of home automation, so here the focus moves to what changed later.
Several limits held this period back. Devices from different makers did not work together. Setting up a system often needed a specialist. Feedback was poor, meaning a controller often could not tell whether a lamp had actually turned on.
Wireless protocols in the 2000s
The next shift in smart home networking came with low-power wireless networks designed for sensors and switches. In the early 2000s, new protocols appeared that let small battery-powered devices talk to each other through a mesh, passing messages along from one device to the next. Z-Wave and Zigbee are the best-known examples, and Bluetooth and Wi-Fi gradually took on similar roles. The advantage was that a switch, a motion sensor or a door lock could be added without cutting into walls.
At the same time, microcontrollers and radio chips became cheap, a development tied to the spread of silicon chips in every kind of product. Sensors could be added to ordinary objects, which fed the broader idea of connected sensors in homes, cars and factories.
Learning and app-controlled thermostats
Around 2007, companies began selling thermostats with touchscreens and Wi-Fi, which could be adjusted from a computer and later from a phone. The Nest Learning Thermostat, introduced in 2011, drew particular attention. It was designed to build a schedule by watching how people adjusted it, and to detect when a house was empty. Claims about how much energy such devices save vary, and results depend on the household, the climate and the heating system, so any figure should be read with caution. Independent studies have reported modest savings in some conditions.
The smartphone mattered as much as the device. Before phones, every smart product needed its own control panel. After phones with touchscreen interfaces spread, a company could ship a small, cheap device and offer a free app as the interface. This lowered the cost of making products such as connected light bulbs, plugs, cameras, video doorbells and door locks, and it moved the user’s attention from the wall to the pocket.
Voice control
The next layer was voice. Smart speakers from the mid-2010s added a microphone and an assistant to the home, so that a person could say a command rather than open an app. The technology behind them is covered in the account of voice assistants. For the smart home, voice control had a practical benefit: it could act as a single remote control for many devices from different makers, as long as each supported the assistant.
This also brought familiar debates about privacy. Microphones and cameras in the home raised questions about what data was stored, who could access it and how long it was kept, and these questions have not gone away.
Standards and fragmentation
As products multiplied, so did incompatibilities. A light bulb could work with one assistant and not another, or need its own hub. Industry groups responded by developing shared standards. One of them, called Matter, was launched in the early 2020s with the aim of letting devices from different makers cooperate across ecosystems. How completely it succeeds will depend on how widely manufacturers support it, and early results were mixed.
Alongside standards, there is a quieter issue of longevity. A connected device depends on software and often on a company’s servers. If a maker stops supporting a product, features can stop working long before the hardware fails. A mechanical light switch has no such risk.
What stays the same
Behind the new terminology, the pattern is familiar. A sensor measures something, a controller decides, and an actuator switches something. A thermostat from the 1880s did this with a bimetal strip and a relay. A current smart thermostat does it with a processor, a radio and a cloud service. What changed is the number of things that can be sensed and controlled, the ease of setting them up, and the number of places the system can be reached from, which makes the modern version more capable and also more complicated to secure and maintain.