Humanoid Robots: A Reality Check on Deployment

Fewer than 20 companies will deploy humanoid robots at scale by 2028—despite the frenzy of investment and development surrounding them.

This stark prediction comes from Gartner Inc., which reveals a sobering reality for the humanoid robotics sector: while nearly 200 companies are experimenting with humanoid robots, most will remain in the lab far longer than expected. The promise of versatile, cost-effective robots that can navigate real-world environments is compelling but, as Gartner notes, the technology remains immature and often fails to meet the high expectations set by initial demonstrations.

The current landscape is littered with examples of humanoid robots that, despite impressive demo reels, struggle in real-world applications. Take Boston Dynamics' Atlas, which boasts 28 degrees of freedom and has captivated audiences with its parkour-like feats. While the robot can perform complex movements, it remains primarily a research project rather than a field-ready solution. Technical specifications show that Atlas is capable of limited payload capacities and has a runtime constrained by its battery size, making it impractical for sustained operations outside controlled environments.

Gartner's data also highlights a critical insight: only a handful of companies will transition from proof of concept to actual deployment within supply chain and manufacturing contexts. The challenges are significant. Humanoid robots are designed to mirror human movements and capabilities, but this often leads to mechanical complexity that results in higher costs and lower reliability. As Abdil Tunca, a senior principal analyst at Gartner, explains, Chief Supply Chain Officers (CSCOs) must tread carefully, weighing the potential benefits against the current technological limitations.

One key limitation is the robots' ability to adapt to dynamic environments. While they may excel in controlled settings, such as a factory floor, they often struggle when faced with unpredictable variables—like a fallen object or a change in task requirements. This inflexibility limits their applicability in high-throughput operations, where efficiency and adaptability are crucial.

Moreover, as these companies vie for market share, the risk of vaporware looms large. Many claim significant advancements but deliver little more than polished presentations. For example, Tesla's Optimus robot, which is positioned as a general-purpose humanoid, has yet to demonstrate capabilities that would justify its ambitious promises. With a projected power source and runtime still untested in practical application, Optimus embodies the tension between hype and reality that currently characterizes the sector.

In contrast, companies like Apptronik are making strides with their Apollo robot, which features a functional design that emphasizes ease of use and integration into existing workflows. With a reported payload capacity and a focus on modularity, Apollo may be better positioned to meet current industry needs, though it still lacks the robust performance metrics to fully support mass deployment.

As we look to 2028, the takeaway is clear: while humanoid robots may one day transform industries, the path to widespread adoption is fraught with challenges. The technology must evolve to demonstrate that it can deliver on its promises of efficiency and versatility. For now, the reality is that most humanoid robots will remain confined to laboratory settings, as companies grapple with the complexities of bringing these machines into the real world.