A new method for training factory robots could significantly transform military manufacturing, particularly in the development of drones and weapons, allowing for high-volume production near front lines. This advancement highlights a pivotal aspect of the competition between the United States and China in the realm of manufacturing, termed the “pacing challenge” in the National Defense Strategy.
The research, detailed in a paper from the January issue of the International Journal of Extreme Manufacturing, introduces the concept of AI-driven additive manufacturing (AAM). Currently, factory robots are limited to performing a set number of rigid movements, are hard to reconfigure for different tasks, and require specialized setups on production floors. They lack the ability to recognize misalignments or errors during operation.
An international collaboration comprising researchers from California State University, Northridge; the National University of Singapore; NASA’s Jet Propulsion Laboratory; and the University of Wisconsin-Madison has developed a new system. This system trains robots to perform a wider array of human-like movements, enabling a basic level of perception and comprehension. Coupled with 3D printing technology, this framework could facilitate comprehensive manufacturing of electronics, including drones that are essential in the Ukrainian conflict.
This innovation aims to bolster U.S. manufacturing strength, a critical concern amid the ongoing power struggle with China. Before President Donald Trump prioritized re-shoring manufacturing during his presidency, the Biden administration had already initiated various policy measures to achieve the same objective, employing tariffs alongside grants and incentives.
The Pentagon has taken steps to enhance the production of critical components such as microelectronics, recognizing China’s dominance in these areas as a strategic vulnerability. The CHIPS and Science Act of 2022 aims to allocate over $52 billion for domestic semiconductor manufacturing, underscoring a bipartisan acknowledgment of the importance of industrial capability for both military and economic power. A failure to secure domestic access to key technologies could lead to operational delays and weakened deterrent capabilities in potential future conflicts.
However, a shortage of skilled workers poses a significant challenge for fulfilling these manufacturing objectives. Institutions like the National Institute of Standards and Technology (NIST) and industry leaders predict that it will take at least seven years to develop this workforce, even with government investments.
Commerce Secretary Howard Lutnik noted that robotics and automation could play a pivotal role in revitalizing U.S. manufacturing, leading several American robotics firms to advocate for a national strategy to encourage manufacturing automation. Despite this, the current state of factory robotics is still lacking in adaptability and dexterity compared to human workers, as emphasized in a 2019 Boston Consulting Group report.
The newly proposed AAM model integrates autonomy across various manufacturing processes, managing everything from design preparation to post-print machining through a collaborative fleet of robots or drones. This integration is particularly crucial for future manufacturing endeavors in outer space, justifying NASA’s involvement and interest in the project.
The AAM relies on a sensor-integrated design comprising a layered intelligence model, featuring a “knowledge layer,” “generative solution layer,” “operational layer,” and “cognitive layer” that bestow robots with artificial intelligence capabilities. The AI agents act as high-level controllers that assess and utilize the skill pool to optimize task execution and adapt to new challenges.
Co-author Bingbing Li of California State University, Northridge, emphasized the goal of minimizing human operation in the manufacturing process, aiming for higher degrees of automation. Although human experts will remain crucial, their roles will shift from operators to supervisors who guide and enhance robotic capabilities.
The AI’s potential extends to the military domain, suggesting that soldiers could use AI systems to redesign and 3D print replacement parts for malfunctioning equipment in the field, streamlining logistics and manpower requirements. The Department of Defense, including DARPA, increasingly prioritizes secure and scalable manufacturing methods for challenging logistical environments.
Looking ahead, this manufacturing model could offer a roadmap for not only more agile production but also for reestablishing manufacturing in the U.S. This focus has significant implications in relation to China’s industrial strategy, where the country leverages its manufacturing capabilities not just for economic gain but also to build significant military resources.
Former Undersecretary of Defense Bill LaPlante highlighted the urgency of this issue in October 2023, noting that the U.S. is in a critical technological and economic competition to maintain its edge in manufacturing crucial defense systems. Nevertheless, the challenge remains, as China currently leads the market share in essential robot components such as sensors and actuators, alongside greater access to materials necessary for automated manufacturing processes.
While robots have the potential to enhance U.S. manufacturing capacity, it is paramount to recognize that they cannot build themselves.
