Humanoid robots are moving beyond the silver screen and into factories, hospitals and logistics hubs — though you’d be forgiven for missing the revolution. There’s no dramatic Terminator entrance, no Ex Machina mind games. Just the quiet hum of hundreds of factory robots working in darkness while their human supervisors sleep.
According to the Institute of Electrical and Electronics Engineers’ (IEEE) global study The Impact of Technology in 2025 and Beyond, 18% of technology leaders expect full humanoid integration in their organizations this year, while 37% are considering implementation.
Current deployments span the mundane to the critical: robots scrubbing hospital floors, performing hazardous inspections in plants, greeting customers. But what separates this generation from their predecessors isn’t the task list. It’s the cognitive leap. These machines now demonstrate advanced autonomy, adaptation and the kind of dexterity that makes folding a fitted sheet look effortless.
Ready for the factory floor
“Most companies have already used AI to test specific areas,” says Christian Pedersen, chief product officer at industrial AI firm IFS. “Adoption accelerates when AI solves things that are dirty, dangerous or difficult — because people actually want to use it.”
Pedersen describes the emergence of what he calls “physical AI”: intelligent systems that don’t just analyze data, they physically interact with the world.
A recent partnership between IFS and robotics company 1X Technologies emphasizes how quickly this market is developing. The companies are combining 1X’s humanoid robots with the IFS.ai platform to build production-ready robotics for manufacturing, utilities, aviation and other industrial sectors.
Meanwhile, 1X’s humanoid robot NEO is already moving into homes, with preorders open in the US. The robot can tidy rooms, water plants and fold clothes — tasks that sound mundane but represent a major technical leap.
“Folding clothes is extremely difficult for robots,” Pedersen notes. “But that kind of dexterity transfers directly into industrial environments.”
This mechanical dexterity represents one half of the equation. The other half — sensory intelligence — is advancing just as rapidly. Boston Dynamics’ (which has also recently announced it is using IFS.ai) robot dog Spot can detect steam leaks in a factory by their sound alone. “That can pay for itself in a month,” he says.
The cybersecurity challenge
But as physical AI accelerates, so do the risks.
The IEEE study shows that 41% of technologists expect their organizations to begin implementing robotics-specific cybersecurity solutions this year, with another 33% planning full implementation. These systems use AI to monitor robots in real time, identify threats and prevent data leaks or financial losses.
This urgency to address cybersecurity risks, Pedersen says, is justified. “Security in the AI era has many layers,” he says. “If you manipulate the data feeding an AI, you can cause serious harm.”
He is particularly concerned about general-purpose AI agents that lack enterprise-grade security controls. Pederson references a cybersecurity conference where researchers queried a customer support AI system and, through carefully crafted prompts, extracted the organization’s entire customer list.
“Proofs of concept often work in isolation,” he says. “But when you try to deploy them in the real world, that’s where things break.” With this, IFS has built security and access controls directly into its agentic AI systems, says Pederson.
China’s ‘dark factory’ revolution
Globally, humanoid deployments are already accelerating beyond pilot projects. Mohamed Kande, global chairman of PwC recently said on stage at IFS’ conference that he was “blown away” by the scale of humanoid robots in automotive plants in China — some operating as “dark factories” with no lighting or air conditioning, delivering major energy savings.
What makes this wave different from earlier automation, Pedersen argues, is adaptability. “There are already robots stacking items 30 feet high in warehouses,” he says. “But humanoids can take on more nuanced, intelligent tasks.”
Over the next year, he expects to see real-world case studies of humanoid robots becoming standard coworkers in industrial settings. “Humanoids are no longer just about movement — they’re about judgement,” he says.
