(CNN) Picture a robot.
Did you envision a giant machine assembling cars, Data from "Star Trek," C-3PO from "Star Wars" or "The Terminator"? Most of us would probably think of something massive -- or at least human size.
But a whole arm of robotics is focusing on bug-size 'bots (and smaller).
It's not just the size of tiny insects that are inspiring roboticists; it's also the many complex tasks and physical feats that comprise the everyday lives of many fleas, flies and other six-legged creatures.
The question isn't only how big and powerful we can make a machine, but how small and savvy. What might humans be capable of if we could command a tiny army of simple machines? How could we use robots that could fly, skim across the water, hop to the ceiling and even swarm?
That's exactly the kind of question roboticists like Zeynep Temel, an assistant professor of robotics at Carnegie Mellon University in Pittsburgh, is asking -- and answering -- in her lab.
Tiny robots would be useful in medical applications -- for targeted drug delivery or simple surgeries sans incisions, according to Temel. Miniature robots could also save lives in dangerous places like minefields, or during search and rescue: "If you have small bug-robots," she said, it's possible to do "more efficient -- and safer -- rescue operations" following an avalanche or earthquake where it's dangerous for humans or even larger robots to tread.
Small robots that can work together, like ants or bees do, would also be ideal for exploring other planets like Mars, again keeping humans away from risky, unexplored situations:
"I hope my research will be used to make modular robots that can self-assemble, to be used by astronauts in unknown environments to lend a helping hand," said Jamie Paik, founder and director of the Reconfigurable Robotics Lab at the Swiss Federal Institute of Technology.
These are just some of the important applications bio-inspired robots could be used for, and that's why roboticists at the worldwide major robotics labs are dedicated to exploring the class Insecta.
Ants are a favorite inspiration -- they're able to lift bulky and heavier-than-they-are loads and travel quickly in sandy deserts as well as woodlands. These insects also work together to create bridges and surmount obstacles.
Trap-jaw ants served as a model for a team that developed a battery-powered, palm-size robot that can "adapt to an environment and can collaborate," Paik, a member of the team, said.
In the natural world, trap-jaw ants do all the ant things — and they can also snap their powerful jaws at the incredible speed of 90 miles per hour to jump away from predators. Paik and her team used the same mechanics to help give the robots a variety of movements, including "vertical jumping for height, horizontal jumping for distance, somersault jumping to clear obstacles, walking on textured terrain and crawling on flat surfaces," according to the paper's abstract.
Like the ants they are based on, each unit is completely autonomous, but they can communicate and thereby work together via a simple transmitter.
Another advantage to tiny, fairly simple autonomous robots? They're cheap, compared to a larger robot. "We can throw out multiples of them, and if we lose or break some, they can still execute a given task," Paik said.
Why are insects such useful inspiration for roboticists? They give scientists a starting point, Paik said, that proves what's possible -- such as jumping 100 times their height like a flea, climbing vertical obstacles (or even upside down) or packing full-size wings under a petite hard shell like a ladybug. "These are nature's optimized designs," she said.
Temel backed up Paik's point -- while she said she's still a bit personally afraid of insects (the live ones), she has nonetheless come to admire how well they solve so many difficult problems.
"They swim, they fly fast and know how to balance to perch, they walk and jump on the water's surface, and they can jump onto leaves which are like tiny, unstable platforms," Temel said.
"All these movements and motions that we are still trying to do in large-scale robotics and (insects) are doing it all -- and they are packing it all into such a small space."
One of the biggest challenges for robots of all sizes is remaining upright on complex surfaces. "Robots are good at smooth, but natural terrain is fascinatingly rough, and so it's difficult for robots," said Kathryn Daltorio, an assistant professor of mechanical engineering at Case Western Reserve University.
Insects have six legs -- or two sets of tripods, thinking robotically -- which are great for stability. In addition to the multiple legs, they also utilize a variety of structures and materials to move across vertical surfaces, which Daltorio has studied and mimicked to create her very simple but successful climbing Mini-Whegs, which only have one motor.
One thing humans can't replicate exactly (yet) are the unique materials grown or excreted by insects. Tiny claws, spines and sticky pads allow insects to perform their many feats.
Daltorio was able to use human-made materials like Velcro, Scotch tape and craft spines on her Mini-Whegs, but other materials, like the lightweight, biodegradable hard exoskeletons are tougher to reproduce. For now, plastics and carbon fiber can serve to make inexpensive, easily produced exteriors, according to Daltorio.
But not all insects are hard, and roboticists are now creating soft robots that can change shape to enable movement like caterpillars can, instead of relying on legs for every type of movement. A shape-memory alloy can be used that reacts to temperature, so a soft robot can curl and roll but "remember" a stiffer shape when heated. "You can bend them, and apply a heat source, and they will bend back — we use those a lot," Temel said.
Electricity can also be used to change the shapes of special piezoceramic materials. "We apply voltage to bend them one direction or another, which is useful for cockroach-inspired robots," she says.
Speaking of power, that's another challenge for tiny robots. Rechargeable batteries and solar panels are two popular solutions. The RoBeetle utilizes methanol to contract and expand its lab-designed "muscles."
But like the materials challenges, this is another area where roboticists work closely with colleagues who specialize in materials science, battery and solar technology to work toward autonomous robots. Building a robot is a collaborative endeavor, Temel said.
There is still a lot more research for roboticists to do before robots are as smart and agile as insects: "Even a small fly has really smart reactions that are quick and able to respond to a lot of sensory information," Daltorio said. "(Insects) have this very rich and adaptive behavioral repertoire that we can't yet get a robot to do."