For the first time, engineers have cracked the code on providing robots with complex instructions without relying on electricity, paving the way for a significant boost in robotic cognitive capacity.
Scientists at King’s College London have successfully transmitted a series of commands to devices using a novel compact circuit that mimics the human body’s physiology by harnessing pressure changes from a fluid within it.
This revolutionary technology introduces the prospect of autonomous robots, capable of functioning independently from their central control systems, paving the way for more sophisticated AI applications in place of traditional software solutions.
By offloading tasks to distinct bodily parts, robots can liberate processing power to “think,” paving the way for more socially aware or agile future generations. According to Dr. Antonio Forte, a Senior Lecturer in Engineering at King’s College London and lead researcher, this breakthrough has the potential to introduce a novel breed of robotics that can revolutionize fields like social care and manufacturing.
The research yields results that enable the development of robots capable of operating in environments where traditional electricity-powered devices are ineffective, such as radiation-drenched areas like Chernobyl or sensitive electromagnetic spaces like MRI rooms, where conventional systems would be destroyed.
Researchers also anticipate that these robots could ultimately be deployed in areas with limited access to reliable electricity, specifically in low-income countries.
“Essentially, robots can be categorised into two fundamental parts: their cognitive or ‘mind’ component and their physical structure.” Despite advancements in artificial intelligence, many robots still struggle to accomplish even the most basic tasks, such as opening a door – a seemingly straightforward challenge that has left experts perplexed.
Software development has accelerated rapidly recently, but hardware hasn’t kept pace. By decoupling {hardware} systems from the software running on them, we can effectively shift a significant portion of the computational burden to the {hardware}, just as your brain doesn’t need to instruct your heart to beat.
Currently, all robots rely on electrical power and microprocessors to function effectively. A robotic ‘mind’ comprising complex algorithms and sophisticated software interprets vast amounts of information and transmits it to the physical hardware via an encoder, ultimately executing a specific action.
Tackling the complexities of “tender robotics,” where innovative applications such as robotic muscle tissue are crafted from sentimental materials, poses a significant challenge. This involves overcoming hurdles like integrating arduous digital encoders and placing demands on software that requires sophisticated fabric behavior, for instance, grabbing a door deal with.
To overcome this challenge, the team created a reconfigurable circuit equipped with an adjustable valve that can be seamlessly integrated within a robot’s hardware. This valve operates similarly to a transistor within a conventional circuit, enabling engineers to transmit commands to {hardware} using pressure, which effectively encodes binary data, thereby allowing the robot to execute complex movements without reliance on electrical power or instructions from its central control system. This design enables more effective management compared to traditional fluid-based systems.
By transferring the workload from the software program to the hardware, the innovative circuit creates a window of opportunity for next-generation robots to become increasingly sophisticated, adaptable, and valuable in their applications.
Subsequent to these experiments, the research team aims to upscale their circuits by integrating them into larger robots, including crawling machines that can monitor energy-rich flora and wheeled bots equipped with delicate propulsion systems.
Without further funding, the development of embodied intelligence in robots is likely to plateau. If we fail to offload the computationally intensive tasks that modern-day robots typically handle, any algorithmic advancements are unlikely to yield significant improvements in their overall performance. Our work represents just the beginning of this journey, but the future promises even more advanced robots paired with innovative human physiology.
