When you hear the word “robot,” what comes to mind? Most of us would first picture smaller devices like a Roomba or drone, or maybe the iconic R2-D2. But what if we imagine more complex robots? Do you think of the cars we take to work, from a 1998 Subaru Outback to a Waymo or Tesla? What if we went even bigger, like airplanes on autopilot or some of the science fiction (maybe soon to be reality) androids, humanoids, and Transformers—whose tagline is literally “robots in disguise”—commonly portrayed?

So what actually constitutes a robot? Size isn’t a limiting factor; it’s not how it looks or whether it travels. There’s a wide variety of forms, from a stationary robotic arm making microchips, to a deep-sea submersible mapping the ocean, to a firefighting drone suppressing a wildfire. Robots are all around us, and we spend more time with them than we might think.

What’s the definition of a “robot”?

A robot is a series of connected systems that’s capable of executing on any or all of three things: sense, compute, and act. It’s also capable of performing a range of mechanical tasks without direct supervision based on feedback from its environment. Essentially there's some aspect of “autonomy,” even if only at a low level—running a series of “if-then” control loops.

Robots in disguise

The examples so far are ones you may not have personally interacted with, but are at least vaguely familiar. However, there’s a massive but overlooked population of robots we interact with every day.

Today you probably woke up in a home that’s heated or cooled automatically or walked past sensors for doors, lighting, or security at a factory, school, or gym. You might work in an office or shop in a supermarket that’s climate-controlled for comfort and food preservation, respectively. We are constantly in and out of buildings—we spend most of our lives in them—and in order to keep these indoor places hospitable, we depend on robotic systems. Does this make buildings…robots?

We say yes. Buildings, like Transformers, are “robots in disguise.” These infrastructural robots are actually the largest and most complex robots by orders of magnitude, with sensors, computing power, and control points (temperature, airflow, lighting, etc.) that provide comfort and safety.

While a self-driving car may have dozens of sensors, they have only three major control points: steering, braking, and acceleration. By contrast, some of the largest buildings are trending toward 1 million I/O sensors and controllables! Even a small building can have hundreds of I/O points, including the growing number of IoT devices we find inside. That a very complex robotics system!

Why does reframing buildings as “robots” matter?

For too long, our vision of what constitutes a robot has been too narrow. We’ve focused on making smaller robots—cars, buses, and tractors—more energy efficient but have wholly overlooked an enormous opportunity for energy reduction among the largest robotics. The complexity of buildings’ controlled systems makes them an excellent opportunity for AI technology to better manage the systems.

Buildings have greater potential for energy savings in the world economy than either industry or transportation. Yet right now, building control systems are mostly reactive, have a little-to-no level of self-awareness, and have little ability to autonomously learn and make smarter control decisions in response to their environment. Thus, they are nowhere near as efficient as they could be.

Modern building systems are not forward-looking or proactive. They can’t anticipate or adjust controls in advance to account for things like changing weather or grid needs. But before we can make dramatic and necessary efficiency improvements, we all need to reframe our mindset and agree on unified terminology and definitions so we can work toward and evaluate success.

Though realigning expectations seems daunting, it’s been done before. For decades, there was no major innovation in the automotive industry. But reframing how the industry thought about the levels of autonomy for vehicles allowed for major advancements in the field, leading to the development of the aforementioned Teslas and Waymos.

By recognizing buildings as robots and defining a similar scale, we can drive more attention and investment toward the built environment. Making buildings truly autonomous will improve comfort, increase sustainability, save time and money, and move us toward the benefits of truly smart cities.

Defining levels of autonomy for buildings

What makes a building “smart” or autonomous? As we’ve noted before, we have smart phones, smart appliances, and more, but we lack consensus on what “smart” means, especially in the built environment and how to measure the ability to act autonomously.

In 2014, the Society of Automotive Engineers (SAE) established a taxonomy for vehicle automation. Level 0 includes basic cruise control, while Level 1 has adaptive cruise control—braking and adjusting based on the position of other vehicles. SAE considers the pinnacle to be Level 5, when a vehicle has the ability to drive itself anywhere and in all conditions. Rallying around common goals and definitions greatly benefitted the autonomous vehicle industry and drove investment and innovation.

We've taken inspiration from the SAE taxonomy and developed the Levels of Autonomy for infrastructural robots below. Our vision includes levels of autonomy beyond 5, when a building functions as a partner with us and acts as an independent agent.

For instance, a truly autonomous building at Levels 6-8 will guide an installer through installation, adaptively program itself every day, and continuously self-commission and validate its systems. It could negotiate with other buildings nearby or utility companies, buy and sell energy futures, and cooperate peer-to-peer in smart, interactive cities.

Improving comfort in a changing world

Autonomous buildings move controls from a reactive paradigm to a proactive one.

We need more sophisticated, AI-driven advancements that keep up with our changing world. As the global climate shifts and energy demands become more variable, our places of shelter must remain livable while maximizing efficiency.

Instead of simply responding to fluctuating temperatures, an autonomous building could anticipate conditions based on the local climate, sun path, weather forecasts, and self-knowledge of its systems. It could start cooling early for a hot forecast, draw blinds automatically as the sun moves, or change ventilation based on occupancy.

Autonomous buildings would know how to prioritize human comfort while adjusting for the type of activities held inside, from data centers, to coffee shops, to yoga studios, to manufacturing warehouses.

Increasing sustainability in the built environment and beyond

Buildings are the largest controlled infrastructure in the world. They are responsible for 30-40 percent of all energy consumption and emissions worldwide and in the U.S. Households and commercial buildings in the U.S. consume more energy than the transportation or industry sectors. Remarkably, 30 percent of building energy is wasted due to poor system control and related factors, representing the largest opportunity in the economy for cost savings and emissions reduction.

Improving robotic infrastructure in our homes and workplaces would dramatically enhance our ability to conserve energy while simultaneously providing more comfortable places to live and work.

Saving time and money with smart buildings

As of May 2024, there are over 21,000 vacancies in the Building Automation Systems (BAS) profession, according to the Association of Controls Professionals. There’s a severe shortage of this type of labor, and overall demand for technical skills is expected to rise 55 percent  by 2030. Robotic infrastructure would save time and money by helping us do our jobs more efficiently. For example, a smart building would identify its own maintenance needs and initiate a request, like an error code generated by a vehicle.

An autonomous building that knows how to reduce its own energy consumption would be more economical, fine-tuning its own sequences to optimize the system before it gets too hot or too cold.

Infrastructural robots and smart cities

Imagine a world where the building you’re in isn’t just a passive structure but a proactive partner. A space that anticipates your needs, adjusts to your preferences, and even contributes to a more sustainable future. This vision isn't science fiction; it's the future of the built environment.

In a world concerned about climate change, truly autonomous buildings and cities would save energy, coordinate power consumption across the grid, and prevent blackouts during peak usage. Technology developed for use in autonomous buildings could also power other infrastructural robotics that would improve our lives, such as improving manufacturing, logistics, and supply chain facilities.

Our buildings already have some of the pieces in place to enable this autonomous future. It’s time we reframe buildings as infrastructural robots, properly recognizing their potential to reap the benefits of comfort and efficiency.