When Can We Expect to See Human-Like Robots in the Workplace?
The potential of humanoid robots is enormous, with Citibank and Morgan Stanley estimating the market could be worth trillions of dollars by 2050 – but how close is this to reality?
- According to Wall Street predictions, the humanoid robot market could become a multi-trillion-dollar industry over the next few decades.
Robotic co-workers may be closer than you think. Around the world, businesses are dabbling with human-like robots designed to work alongside colleagues in factories, offices and other workplaces, bringing down the cost of labor and improving accuracy and efficiency.
Much of that innovation is coming from China, where 52 percent of all industrial robots installed in 2022 were located, according to the Information Technology & Innovation Foundation. But in the U.S. and Europe, a slew of companies are working to design, develop and deploy human-like robots to operate alongside us in the workplace and home.
“New innovations are enabling capabilities that we couldn’t have imagined only a few years ago,” says Neil Shelton, Senior Vice President of Strategy and Investments at GXO, a logistics company that is going all-in on human-like robots.
The marketplace is moving fast. Humanoids are already on factory floors doing tedious grunt work that people do for minimum wages and sore muscles. The next step could be robots in the home, where they may be able to take over many household chores, freeing up people’s time while providing a godsend for countries with ageing populations and ever-shrinking numbers of offspring available to look after elderly relatives.
The economics behind this are nearly here. Investment bank Morgan Stanley calculated that a $50,000 robot saves the user at least $500,000 over 20 years compared to a low-wage worker. So far, costs are higher than that but they are coming down with every new advancement. The end result is estimated to be a total addressable market that could end somewhere in the multi-trillions of dollars, according to some Wall Street predictions.
“New innovations are enabling capabilities that we couldn’t have imagined only a few years ago”
At GXO, robots are being used to perform repetitive tasks, such as engraving products, programming electronic devices, and building pallets in customized configurations, with high precision, says Shelton.
But it’s the humanoid robots that have gained the most media attention – and could prove the most transformative. “Robotic systems, such as humanoids, reduce labor-intensive activity, increase efficiency and safety while freeing up team members to focus on higher-value-add activities,” Shelton explains.
Robots can be split into three types, explains Jonathan Aitken, Senior University Teacher in Robotics at the University of Sheffield. Industrial robots tend to be little more than arms, performing repetitive tasks in manufacturing facilities, such as putting together cars at scale. Personal service robots perform single-skill tasks in the home – think of a robotic vacuum cleaner, for instance, or a lawnmower that trundles happily over your grass while you’re doing something else.
But it’s the third type of robot where human-presenting machines hold the most promise: professional service robots. These are robots that perform more general jobs in the workplace and are usually mobile, explains Aitken. Traditionally, they’ve not looked much like humans – he highlights Amazon’s Proteus robot, a low-to-the-ground wheeled vehicle that picks up and moves containers full of items in Amazon warehouses.
Currently, Proteus operates in specific spaces, and Amazon clearly delineates areas where its human and robotic workers operate. The next step is for robots to work on the same production lines in a complicated dance with human colleagues.
Therein lies the next challenge for robot development. “The key to using robots within different areas is unlocking the potential for robots to operate in more uncertain or unknown environments, especially ones where people will be present,” says Aitken.
GXO is on the cusp of that shift towards more human-like robots, and has been training them using artificial intelligence to improve efficiency.
“The defining factor for a successful robot will be how well it adapts, either to its process or where it finds itself”
The AI revolution, kickstarted by the 2022 release of OpenAI’s ChatGPT, is positively impacting the evolution and adoption of more human-like robots. That process is being further accelerated by the work of companies like EQT Ventures portfolio firm, 1X, a robotics designer, which in September 2024 unveiled a generative AI model that could predict how robots would interact with their world, making them more like humans.
1X has already developed a wheeled robot for business use, called Eve, and is developing a bipedal, humanoid robot, Neo, powered by generative AI to interact with the constantly changing environment of a busy domestic household.
The 1X generative AI model is “a radically new approach to evaluation of general-purpose robots”, Jack Monas and Eric Jang of 1X wrote in a blog post. Nvidia has also developed a foundational AI model specifically for robots, called Gr00t, which it hopes will help the adoption of human robots in workplaces.
Those AI world models will likely be vital to help further advance uptake of human-like robots in workplaces, which can often be unpredictable by their nature .
Similar work is being done by Vsim, a UK-based company also backed by EQT Ventures, which develops simulation technology and software solutions for robotics that help AI agents solve complex tasks.
Once those robots are equipped to enter the free-for-all of a workplace, they could have a huge effect on the future of work. The investment being put into the space by EQT Ventures and others should help bring down the cost of developing and deploying human robots into the workplace.
But before that future can be realized, further work is required to tackle the tough challenges of robotics. After all, no business or family is going to hire – or retain – a robot that gets in the way and clashes noisily around an environment. “The defining factor for a successful robot will be how well it adapts, either to its process or where it finds itself,” says Aitken.
ThinQ is the must-bookmark publication for the thinking investor.
