Humanoid robots in the sheet metal industry: who do they help and when can they be used?

Humanoid robots are no longer science fiction. While companies such as Boston Dynamics, Hanson Robotics, Universal Robots, Apptronik, Agility Robotics, and Figure.AI continue to drive technological progress, car manufacturers like Tesla, Toyota, and BMW are also entering this market now. They either develop their own robotic solutions or prominently showcase humanoid robots in their production facilities.

Nevertheless, many companies remain skeptical about using robotics in production. Existing robotic solutions are often  seen as expensive, inflexible, and difficult to integrate into existing production processes. Particularly in metal processing and sheet metal manufacturing, conventional industrial robots require significant efforts in terms of operation and programming. As a result, many businesses continue to rely on human labor instead of automated solutions.

At the same time, the demand for flexible production is increasing - for small batch sizes as well as hybrid production scenarios where humans and machines collaborate closely. Traditional automation frequently reaches its limits under these conditions.

The promise of humanoid robots lies precisely in their potential for greater flexibility compared to conventional industrial robots, as they can perform human-like movements and easily integrate into existing workflows.
 

Rapid developments in artificial intelligence are enabling robots to learn new tasks independently. At the same time, the cost of hardware and software is falling rapidly.

The vision is compelling: manufacturers might soon deliver their machine tools directly equipped with an integrated operator - a humanoid robot. These robots could collaborate with humans during the day and autonomously produce large batch sizes at night, potentially transforming production environments characterized by high product variety and variable batch sizes.
 

Despite this promising vision, significant challenges remain. Currently, most humanoid robots are technological demonstrations rather than economically viable solutions.

In particular, interconnected automation solutions clearly hold the advantage in sheet metal manufacturing. These established systems are efficient, cost-effective, and specifically optimized for their applications. As of now, humanoid robots offer no clear advantage in this context.

Nevertheless, humanoid robotics could become highly attractive for flexible, automated small-batch production. The key lies in developing a highly functional robotic gripper - a human-like hand capable of precise, versatile, and sensitive tactile movements. Such a robotic hand could drastically reduce production setup times.

However, this development involves a major challenge: the sensor technology, complex controls, and machine learning required by these robots demand considerable technological and financial investment. Robots must perfectly replicate complex gripping movements, perceive their surroundings in real-time, make autonomous decisions, and flexibly react to unexpected situations. Currently, the necessary computing power, sensor technology, and advanced control systems remain insufficient.

Additionally, the profitability of humanoid robots remains uncertain. While conventional automation solutions rely on established standards and quickly amortize their costs, humanoid robots currently involve substantial development and acquisition expenses.
 

While waiting for this vision to become fully achievable, intermediate solutions are already available. For example, compact and easily positionable bending cells efficiently produce large batch sizes. These cells occupy minimal space, require low programming efforts, and are particularly suited to existing ("brownfield") production environments. Cooperation with modern operating systems from partners such as Intrinsic also facilitates the flexible use of basic actuators (FANUC, Kuka, UR, Comau, etc.).

As AI progresses and interfaces become increasingly intuitive, the effort required for programming these systems will continue to decrease. The target vision: a production environment where humans and machines collaborate flexibly during the day, while the factory operates autonomously at night (Lasting Transformation: why digital sophistication is going to be a competitive advantage | Metal Interface). For example, the positionable bending cell could simply be moved up to the bending machine and integrated into the production process - this approach would be ideal for manufacturing simple products in large batch sizes. Small batch sizes and complex products that require increased programming effort could be produced manually when the positionable bending cell is undocked from the bending machine.
 

Technological development advances inexorably. Continuous progress in machine learning, control systems, and sensor technology is steadily bringing humanoid robots closer to industrial reality.

The dream of humanoid robotics remains fascinating, but it will still require considerable time to become practically viable. Experts estimate a timeframe of 40 to 50 years as realistic. Until then, compact, modular systems represent the pragmatic path towards an intelligent and automated future.