The promise of household robots has remained elusive for decades, but now engineers are thinking about how to build machines that we can trust in our most intimate spaces.

The idea of having a household robot in our homes — that can do the dishes, the laundry, and all the other chores we’d rather not be spending time on — has long been confined to the world of science fiction. But now, as technological progress in AI models, material science and robotics rapidly accelerates, one company believes we’ll see these mechanical helpers in our homes within a decade.

Oslo, and San Francisco based 1X has been working on this problem since launching in 2014, and has developed two humanoid robots — Eve and Neo — where the latest product, NEO, is what they aim to mass-produce and create a society where humans spend less time on mundane tasks, and more time on being human.

And while the engineering challenge of building an affordable robot that’s able to safely move around our houses and handle delicate jobs is a tough one, these machines have also got to win our heart, minds, and trust if they’re to become a commercial success.

So can robots really be designed to work effectively in our homes, while also building empathetic bonds with humans?

1X’s CEO and founder, Bernt Øivind Børnich, shared his vision with Danica Kragic, a Professor of Computer Science at the Royal Institute of Technology, at this year’s EQT Ventures and EQT Growth Scale Summit.

Not just something that replaces your dishwasher

For humans to trust robots that can move around our homes at their own will, we will need to see them as ‘intelligent companions, and not just as something that replaces your dishwasher,’ said Bernt.

To get there, robots will need to be ‘safe, capable, and affordable’ while also being lightweight and soft to the touch — two things that the average robotic systems don’t tend to be.

Unlike some clunkier early attempts at building humanoid robots like Honda’s Asimo, 1X’s Neo has an outer skin of soft fabric. And, rather than opting for the type of lifelike — and arguably creepy — facial design we’ve seen from robots like Hanson Robotics’ Sophia, they have a screen for a face that’s closer to the cutesy design of Eve from Pixar’s Wall-E, after which 1X’s first robot is named.

The softness of 1X’s robots’ appearance is matched by their movements. With the help of powerful proprietary motor technology, they are controlled by artificial ‘tendons,’ which pull against the system’s artificial joints to manipulate its arms and digits, making their movements less jerky — and less likely to break things.

‘It’s very much similar to how human muscle operates, and you get these beautiful dynamics you see in nature,’ Bernt said. ‘It doesn’t actually move like a robot. It moves more like a human.’

This mechanical understanding of how to operate smoothly within our human spaces will be a key step to building understanding between people and robots, and Danica pointed out that engineers now have to build this environmental understanding into their systems from day one.

‘As engineers, we’ve always been thinking, “You build a machine, and then we give it some intelligence.” But here, it’s about thinking about the body and the intelligence as one thing,’ she said.

Intelligence comes from diversity

To direct these human-like movements, 1X’s robots are controlled by advanced AI models developed in partnership with OpenAI, who led a $23.5m investment into the Norwegian startup March, 2023. In 2024, EQT Ventures participated in the Series B $100M funding round along with other notable global investors.

Bernt believes that — to build an empathetic connection with humans — intelligent robots need to be able to predict when they’ll be needed to help around the house.

‘It’s the ability to understand your context and be able to say: “I anticipate that you need help,”’ he said. ‘When we launched this […] almost everyone was like, “Wait a minute. This is completely different, because it’s like there’s a person here.”’

To develop this contextual understanding of their environment and learn when help is needed, household robots need a broad set of training data that prepares them for the various types of situations they might encounter in a home.

‘Intelligence comes from the diversity of our experiences. This has become more and more clear for big multimodal models and text models, getting the best model is not about the amount of data or amount of compute. It’s not about the best architecture. All of these things are pretty similar. It’s about the diversity of your data,’ Bernt said, adding that a robot will learn much more quickly in a home than in a more controlled, factory-style environment.

‘The most diverse environment there is is a home.’

To gather this data, 1X now needs to get its robots into our homes as the company goes from being a pure R&D operation to one that’s more focused on product, and what customers might want. Bernt emphasizes that the first customers who adopt his company’s robots into their homes will be key players in refining that product.

‘What we are launching very soon is the first generation of something that will change how we view humanity and live, but it’ll take us at least a decade to fully realize this. And the first customers that come in here and join us for this, they are pioneers,’ he said.

Danica — who’s extensively researched how robots interact with humans on everyday tasks — compared 1X’s working process with early customer testers to the rollout of robotic vacuum cleaners, which were developed in close dialogue with users.

‘It’s important to have the customer in the loop, because the interaction with the customer will also be a source of information on how we can make the next generation a little bit better,’ she said. ‘Taking that step of bringing it [a household robot] close to untrained humans has been such a big hurdle for many of the companies that we have seen so far.’

Regaining dignity, independence and living life on your own terms

Part of creating empathy between humans and robots will come from the simple fact that they are doing the things we’d rather not.

‘A lot of the attachment that we have to each other also comes not only from our dialog but how we help each other,’ said Bernt. ‘When you have something at home that actually helps you do things that you might not be able to do yourself […] it’s actually a quite magical experience.’

He added that one of 1X’s north stars is to help elderly people, and those with special needs or disabilities, and the company has been doing research into what people in need of care actually want.

“For the people we speak to, it’s about regaining their dignity, independence, and living life on their own terms. They don’t want strangers coming into their apartments.”’ said Bernt.

If household robots can fill this gap, by providing much-needed support to aging populations, there’s a chance that — before too long — we’ll learn to feel grateful that they’re contributing to society.

Bernt asked, “How will we be able to take care of the aging population when there are fewer younger people to assist them? If we can find the balance between receiving thoughtfulness and care from a person, and help from a robot assistant, it would be a crucial problem to solve.”

Danica Kragic, a Professor of Computer Science at the Royal Institute of Technology and Bernt Øivind Børnich, Founder and CEO 1X