Reaching Where Neurosurgeons Can't
Emmett Freeman came to Georgia Tech to build robots small enough to travel through the human brain. The Ph.D. student worked his way toward the problem for years, from a high school research stint on micron-scale drug delivery systems to a run of university labs. When he weighed his doctoral program options, Dr. Azadeh Ansari's lab in TSRB stood out as the most interesting. Ansari, an electrical engineer who directs the MEMS and Micro Robotics Lab, was running exactly the kind of project that piqued his interest: untethered microrobots, built in collaboration with a neurosurgeon at Emory, designed to reach places surgeons’ tools cannot.
That problem sits at the center of neurosurgery. Surgeons work along straight paths, opening the skull and steering traditional instruments toward a target. Deep tumors often sit beyond that reach, and getting to them can damage healthy tissue along the way. “A lot of tumors can't even be accessed because of the limitations with linear trajectories that the surgeons are using now,” Ansari says.
Ansari's lab designs and fabricates tiny devices, the same class of microelectromechanical systems (MEMS) that sit inside phones and computers, then borrows those techniques to build machines that move. Each robot has a footprint of roughly 250 x 250 x 1,250 micrometers (roughly the size of a grain of sand) and travels not by guides or tethers but by external magnetic fields. Because the robots carry no tether, magnets outside the body can steer them along curved paths that a rigid instrument could never follow. The lab runs a wide range of work, from high-frequency communications sensors to electronics that withstand extreme heat, drawing much of its funding from the Department of Defense and DARPA, but the surgical robots most animate Ansari.
The project grew out of a collaboration with Dr. Kimberly Hoang, a neurosurgeon at Emory's Winship Cancer Institute who co-advises Freeman. The partnership began almost by chance. Years earlier, a story on some of Ansari's earliest, non-medical robots went viral, and Hoang saw it. “People in academia tend to live in silos,” Ansari says. “You don't know what's happening next door.” That one connection turned into a project she now counts among her favorites.
Extending the Surgeon's Reach
Freeman frames the platform as a partner to the surgeon, not a replacement. He calls it “an adjunct,” a way of “running a tool for the surgery” that can slip past the blood-brain barrier and reach a target on its own. The robots could carry functional mechanisms to take a biopsy, ferry a dose of medicine to a precise spot, or place an electrode where flexible probes are otherwise hard to steer. A sensor riding along could distinguish a tumor cell from a healthy one by measuring stiffness. Each robot serves a single use and remains sterile for one procedure; the design requirement is for it to complete its task and return.
Ansari sees the same versatility. The technology answers to no single application, she says: “This could be used for biopsy. It could be used for therapeutics or any sort of delivery.” Freeman points to the brainstem, notoriously hard to operate on, as an early proving ground, since the robots could help locate and diagnose them before a surgeon ever cuts.
From Phantom to Patient
“This [work] can help children and adults with neurologic disease and brain cancers that unfortunately still have no good treatments and generally poor prognoses, so the translatable impact for patients is real and needed.”
For now, the work lives in the lab. The team tests its robots in phantoms, gels tuned to mimic the mechanical properties of brain tissue. “We've done all of our experiments on phantoms,” Ansari says. Cadaver tissue comes next, then larger animal models, and eventually living tissue. Ansari expects the regulatory path to stretch on, precisely because the target is the brain. That difficulty is also the appeal. “There are a lot of bottlenecks this could potentially mitigate,” she says.
Building a Company, and a Case
The team is building a business around the technology as the science advances. Ansari, Hoang, and Freeman have filed a patent and gathered early commercialization funding, including a grant from the Georgia Research Alliance and a seed award from Biolocity. Freeman has taken the entrepreneurial side into his own hands, spending the summer in Quadrant-i's 2026 cohort to learn how to turn research into product.
His ambitions for the platform reach past the major hospitals. Freeman imagines it giving rural surgical centers a way to extend what a single surgeon can do. First, the team must prove the idea, robot by robot and tumor by tumor. If it works, Ansari says, the payoff is not one procedure but many, a way to reach the parts of the brain that medicine has long had to leave alone.