Professor Taher Saif has a dream. “It may or may not be possible in my lifetime,” he says. “But it’s the goal.”
His dream is of tiny, semi-organic, intelligent robots—biobots—that will one day swim through our veins under their own power, looking for signs in the blood that indicate cancer. When they detect them, they will track down and attack the tumor cells with a cocktail of proteins, much as our own bodies respond to viruses and bacteria.
It sounds like science fiction, but Saif is already well down the road that could lead to this remarkable idea becoming reality. In his lab at the University of Illinois, he is pioneering research in the relatively new domain of biohybrid robotics—in which artificial and organic parts are combined to form machines.
The traditional image of a robot is something made of metal or plastic and filled with electronic circuitry. But these have drawbacks. They’re heavy, their movements can be rigid, and they can be dangerous around humans. That’s led to the emergence of “soft robotics”, a field which seeks to create more agile, safer machines through the use of non-rigid materials. Biohybrid approaches are at the frontier of that space, and have advanced steadily—underpinned by developments in bioengineering—ever since they were first described in a 2005 research paper.
Right now, biohybrid roboticists are typically concerned with how we can use biological components for the robot’s actuators—the drivers that control how it moves. In conventional robotics, actuators may take the form of motors or pneumatics. In biohybrid robots, these are usually made from muscle cells, and they can be incorporated into systems of a variety of sizes and for a range of capabilities. Alongside a host of potential advantages, from self-repair to energy efficiency, one advantage of bio parts is that they are easier to miniaturize without sacrificing performance.
Saif’s work explores that miniature end of the scale. Together with a group of leading scientists, including his University of Illinois colleague Professor Rashid Bashir and Professor John Rogers from Northwestern University, he has already created biobots the size of a pinhead.
These are made by combining mouse muscle cells with a collagen-based “extracellular matrix” to provide structural and biochemical support. This self-assembles into new muscle, and it is grown around a soft polymer skeleton. Motor neurons, taken from mouse brains or created from stem cells which have been genetically modified to respond to light, are then added. When a couple of biobots are attached either side of a tiny, circular LED activated remotely via wireless microelectronics, the neurons fire and the muscle tissues contract. A series of contractions results in motion: The robots have been designed to either “walk” on a flat surface or “swim” in fluid, the skeleton acting as a kind of spring. Their top speed so far comes in at 0.83mm/second.
But this is just the beginning. The next step will be to adapt the neurons so that they are capable of making decisions for themselves—the reliance of biobots on external controllers is a current limitation. Saif calls this his “mind in vitro” project. “The question is whether neurons on a petri dish can do things that animals do,” he says. “Can they remember things that they have experienced before? Can they develop logic?”
Over the next few years, Saif plans to train his swimming biobots’ neurons, so that they develop a primitive memory, and ultimately be able to make decisions on their direction of travel. Training will be based on punishment and reward. “For example, if a swimmer moves to the right, it may be punished by a chemical, but if they go to the left, maybe they’re given sugar,” he says. “If they are given this choice multiple times, we hope that these neurons would automatically go to the left. They would progress just like we do in our own lives. We make decisions based on the experience we have.”
Which brings us back to Saif’s dream. As well as detecting and treating cancer in the human body, he says, the future biobots could also be used for the detection and clean-up of toxins in the environment and for testing the efficacy of drugs. As he puts it, “We’re paving the way towards intelligent machines.”
Saif’s biobots are tiny for a reason—if tissue grows beyond the scale of a lentil seed, it becomes difficult to supply it with the nutrition and oxygen it needs, and it dies.
At ETH Zurich University, however, researchers are exploring ways to create larger scale biohybrid robots powered by living muscles. Researchers at the university’s Soft Robotics Lab have found a way to 3D “print” muscle cells and gently stretch them into muscle tissue that incorporates inbuilt channels to allow nutrients to be distributed—or “perfused”—inside them, thus mimicking the blood vessel architecture in our muscles.
So far the largest muscle they’ve produced is around two centimeters long and a centimeter thick. According to Dr. Robert Katzschmann, the university’s Assistant Professor of Robotics, the barrier to constructing even bigger muscles is mainly logistical: The two-centimeter muscle alone required around 50 million cells, and a factory-scale facility would be needed to produce cells in large enough numbers for significantly larger muscles. “But in theory, there are no limits,” he says. “As long as you have the cells and you have the perfusability, you can go much larger.”
The lab currently uses so-called “immortalized” cell lines from rats and mice, namely cells that have been artificially manipulated to continue proliferating indefinitely. This attenuates ethical concerns that are instead associated with the use of “primary cells” that have been directly extracted from animals. Katzschmann says that, in theory, new muscle tissues can be built from the cells of any species. “They could come from a grasshopper, basically,” he says. “We just need to be able to expand them to large numbers when cultured in the lab.”
The ultimate goal for Katzschmann is to produce living muscle and tendon constructs that can be integrated with artificial parts to enable the creation of large-scale biohybrid robots. Those muscles can be actuated by electrical pulses. But, much like our own bodies, they will be powered by chemical nutrients. Common nutrients include amino acids and proteins, which are typically supplied in the form of bovine serum, harvested from cow fetuses, which is widely used in cell culture. Possibly—and more ethically—in the future, this will be replaced by something derived from a non-animal source. “It’s a question that hasn’t been answered yet, but we’ll ultimately find ways of producing mixtures of glucose and other nutrients from plant sources that allow the cells to function,” says Katzschmann. Moreover, these bio-actuators can serve as functional models for the fundamental investigation of the pathophysiology of movement, something that is currently carried out mainly via experimental animals.
Katzschmann envisions that these large-scale biohybrid robots could carry out physical tasks currently done by humans. “In principle, they could do anything that needs two pairs of hands and legs if we choose to build them in that way,” he says. Like humans, their biology should afford them dexterity, making it easier for them to enter awkward spaces or adjust to unpredictable terrain. “But we could also build systems that are not so bio-inspired in their looks. They could just be actuators that open a door, for example.”
There are still challenges to overcome. We need to make the processes for biobot creation more scalable and sustainable; we need to create biobots that can integrate a range of functions; we need to extend their lifespan. If these can be surmounted, however, the advantages would not just be about performance, they would also be environmental. Once the biohybrid robots of the future are no longer of any use, they will simply decay. “We don’t need to add to the trash of the world by producing robots that are made of artificial materials,” says Katzschmann. “We will inevitably have robots take over certain tasks for us in the future, so why don't we just build them in the same way that nature does?”
Edward Timpson | Capability Lead & Lead Solutions Architect—Robotics & Autonomous Systems at QinetiQ
Although huge advances have been made in robotics over the past decade, they still have limited purpose within defense. Humans can outperform robots in so many ways on the battlefield, primarily in agility, versatility, and survivability.
However, over the next decade we expect humanoid and bio-inspired robotics to become more mainstream within the sector. These systems, although currently limited by endurance, could give greater agility on complex terrain compared to wheeled and tracked platforms. Biohybrid technology specifically could allow these bots to enjoy more efficient actuation which, when combined with improved battery technology, could open up a multitude of applications.
Furthermore, biohybrid materials that are either self-healing or self-assembling would reduce the need for repair or replacement of components. As we see uncrewed systems becoming the first line of defense in future conflicts, enabling the concept of graceful degradation by which a system can limp back to safety after sustaining damage will be fundamental.
- Mechanical human augmentation. Whether it’s additional limbs or smart exoskeletons, machinery is helping humans upgrade their natural capabilities.
- Power beaming. Sending power wirelessly over long distances could transform everything from electric vehicles to offshore wind farms.
- Neuromorphic computing. Inspired by the brain, neuromorphic chips aim to equal the speed, efficiency, and intelligence of the human mind.
- Gene-editing and enhancement. Advances in biotech are spurring scientists to explore how genomes can be tweaked to make ecosystems more sustainable.
- Hyperspectral imaging. Hyperspectral cameras don’t merely record what something looks like, they can tell you what that thing is made from and help you see what the human eye cannot.
To find out more about QinetiQ, click here
To find out more about WIRED Consulting, click here
