AI is here and it’s creeping into our lives and the way we live. Stuart Corner looks at what it means for the smart home industry and if there are any risks.

Unless you have been living under a rock you would know that artificial intelligence (AI) is all the rage. It’s evolving at a dizzying pace, generating enthusiasm and angst in almost equal measure. No surprise then that AI is having huge impacts on smart home technology and being eagerly adopted and exploited by players in the smart home space.

In 2022 the US-based Association for Smarter Homes and Buildings (ASHB) — then called the Continental Automated Buildings Association (CABA) — published a report entitled The Connected Home and AI, saying it had found thst: “The proliferation of smart systems in connected homes overall, coupled with a new demographics of homeowners who expect a technology-driven experience and the rapid evolution of AI technology, are converging to present a significant new opportunity for connected-home artificial intelligence.”

It said emerging and established participants “should prioritise open ecosystems, where connected home services can leverage multiple devices and disparate data types to provide greater value to homeowners and tenants.”

The report surveyed some 600 homeowners and unit tenants in the US and Canada to ascertain the smart home functions consumers would most like to see enhanced with AI. The list, in priority order, was: home entertainment, home security, heating a cooling systems, lighting and lighting control, energy management and air quality management.

AI will almost certainly make its way into those systems in the not-too-distant future, but the most widespread examples of AI in the smart home today are voice recognition devices and software: Amazon Alexa and Echo, Google Assistant and Apple’s Siri.

One smart home company, US-based Josh.ai, founded in 2015, is an AI company. What Josh, Alexa et al all have in common is that the AI is largely in the voice recognition to interpret human commands to control multiple devices around the home. The decision-making is largely human.

Bringing ChatGPT into the home

Josh’s latest innovation is JoshGPT. It integrates ChatGPT into Josh so that users can ask questions of Josh and have the answers provided by ChatGPT, but this is just another interface to ChatGPT. It’s hardly exploiting AI to make the home smarter.

One real use of AI in the smart home that might not be the first to come to mind is helping humans control and manage what is becoming increasingly complex smart technology. YouTube blogger, Smart Home Solver, has been putting AI through its paces and has been impressed.

In one example he asked AI to program Home Assistant (free and open-source software for home automation) to use a vibration sensor on a door to trigger some actions when someone knocks on the door, but only if the door is closed.

He concluded: “It was really close to figuring out the automation, but it wasn’t quite there. However, this might have been because I didn’t describe the automation perfectly.”

Another application, already well established in industrial settings, is predictive maintenance: using AI to monitor the condition of appliances and notify when they need maintenance or repairs, reducing the risk of breakdowns and extending their lifespan.

This is already starting to happen for individual appliances. Many manufacturers use the Home Connect technology which allows owners, via a smartphone app, to control and monitor appliances that support it.

And individual appliances are already starting to use AI to enhance their functionality. The Home Connect website has a page entitled: Artificial Intelligence in the Kitchen of the Future. It extols the cake-baking capabilities of the Bosch Series 8 oven with AI, announced by Bosch in 2020. “You mix the dough and put it in our oven, press the start button and from then on the baking process is completely automatic.”

Preventive maintenance for people

Preventive appliance maintenance would certainly bring some benefits, but domestic appliances today, by and large, are very reliable and require little maintenance. A much more worthwhile application could be preventive maintenance of people.

Back in 2017 a McKinsey discussion paper Artificial Intelligence the Next Digital Frontier? observed: “AI is enabling the ‘preventive maintenance’ of people, too, and will do even more in the future. Applications powered by artificial intelligence will enable healthcare providers to dramatically accelerate the shift toward personalised preventive medicine. Clinicians will focus on managing patients’ health remotely via wearable wireless sensors, aiming to keep them healthy, fit, and out of hospitals.

“To do this, AI tools will take into account not only patients’ medical histories and genetic makeup but also environmental factors that can influence health, such as pollution and noise where they live and work.”

McKinsey claimed the use of AI to provide preventive care and reduce nonelective hospital admissions could save the UK economy £3.3b (A$6.4b), but it did not specify any timeframe.

Certainly, AI is already being deployed in systems that enable elderly people to live more independently, without intrusive monitoring. Sensors detect indications of normal behaviour, such as lights and domestic appliances being turned on and off and raise an alarm if they detect unusual variations.

A scoping review of artificial intelligence in elderly healthcare, published in January 2023 in Ageing Research Reviews, cast a wide net and concluded: “The roles played by AI technologies in older adults are multiple, and the effect of AI technologies on elderly healthcare is promising. AI technologies are capable of satisfying the unmet care needs of older adults, demonstrating great potential in elderly healthcare.”

The home that knows what your like, and when

Increasingly smart devices that, initially simply provided remote control and monitoring of different aspects of the home in response to human instructions are being enhanced with AI to ascertain normal patterns of behaviour, anticipate needs and self-activate accordingly.

For example, US manufacturer, Teksun, boasts that its thermostats: “Learn automatically from their customers’ behaviour on how to operate and then utilise that information to adjust the temperatures when someone is home or go energy efficient when no one is home.”

This might seem a useful feature, but as smart home technology pervades ever more aspects of domestic existence, becomes ever more intelligent and ever more integrated, the smart home will have very detailed information on many aspects of our lives: what we buy and eat, and when, what we watch and listen to when we come and go.

Just as services like Netflix, Amazon and Spotify make recommendations based on what we have watched, bought, read or listened to, the smart home AI of the future will make recommendations on many more aspects of our lives based on what it has determined are our preferences.

It will know us better than we know ourselves. And that will make smart home AI an attractive target for hackers, and for the ‘good guys’. The Harvard Business Review reports on the hacking of a smart fridge containing “a treasure trove of personal data,” and on the emergence of IoT Forensics: “Snooping around [IoT] devices to find data and, ultimately, clues… secrets that might be invisible to the naked eye. Secrets like when someone switched a light off, brewed a pot of coffee or turned on a TV.”

Such information, the HBR suggests, could be pivotal to the investigation of a crime. However, it could be used, in a totalitarian state, for more sinister purposes.