Wildfires have ravaged parts of Northern and Southern California, in blazes that burned for nearly all of November; tragically taking lives and destroying thousands of homes across the state.
Environmental catastrophes such as these show no signs of waning, as climate change intensifies temperatures across the globe. The fires join a growing number of disaster-related economic losses and fatalities.
But what if a swarm of advanced microrobots could be utilized to eliminate the ‘search’ in search-and-rescue efforts? Such a vision is not merely a pipedream, but a distant reality.
Meet HAMR — Harvard’s Ambulatory Microrobot — a tiny robotic device that can fly, swim, walk on water and race on land. According to Kevin Chen, Forbes 30 Under 30 honoree and one of Harvard’s post-doctorate fellows behind HAMR, the insect-like microbot could transform emergency operations.
“Search-and-rescue missions are usually time-consuming and dangerous for trained personnel,” Chen says, pointing out how a person entering a collapsed site may experience a number of difficulties. “One, it is hard to pass through cluttered spaces; two, his/her motion may cause secondary damage — collapse of nearby structures — and three, other environmental factors may be hazardous, like flooding or a gas leak.”
Approximately the size and weight of a cockroach, HAMR moves at the speed of 25 cm per second and solar-powered versions function for up to five minutes. Not only can it submerge underwater water, but the bot can also zip around quickly on land. Chen, in particular, is responsible for optimizing the mini bot’s cross-functionality capacity, while working toward creating a “swarm” effect. Much like hordes of insects, he envisions the group working together for a common purpose.
Take a wildfire, for instance. If a blaze causes houses and buildings to collapse, a team of hundreds of HAMR microbots could be sent into confined spaces, using cameras and infrared sensors to check for survivors. Chen says the group of automations would collectively scrutinize the area they’re tasked with.
“This can be done in two ways: one, the motion of every robot is controlled by a central computer so each only communicates with a centralized agent/robot, or two, each robot only communicates with nearby neighbors and then makes a decision of where to explore next.” He and four fellow researchers running point on the decade-long project believe the latter is the most promising.
That communication component is crucial to the real-world application. Once a robot sensed a survivor, it would signal to all other bots and pass the whereabouts of the trapped person on to trained personnel nearby.
As it stands, one HAMR costs less than five dollars to make and one week to create. Another area of Chen’s concentration is on improving the robustness of existing models and automating the ways to assemble the microbot. The aim is to commercialize the technology within five years, during which he hopes to cut building time to five human hours per device.
Commercialization won’t automatically lead to large-scale operations, though. The minuscule machine won’t be ready for search-and-rescue deployments for another 10-to-20 years. There’s a lot of work the team needs do before this solution is viable. Power and robustness are the major concerns.
HAMR’s inventors are aware that while solar powering a device is one method toward autonomous machinations, it’s not an all-in-one solution. Limited exposure to sunlight underground would make navigating collapsed buildings for survivors nearly impossible without better circuitry. That’s Chen’s current priority, along with testing durability of the bot weighing 2.8 grams. “Can we make sure the robot survives collisions, and can we ensure the robot is easily fixed so we can use it over and over again?”
A decade or two isn’t too much time, considering what swarms of bug-sized robots could mean for natural disaster relief worldwide. By increasing the efficiency, speed and dexterity of the search-and-rescue system, an indefinite number of lives could be saved.