Drones that dig holes and plant sensors sounds seemingly easy, but it’s not. The NIMBUS Lab at the University of Nebraska has been developing drones that carry a portable digging system, locate a spot to dig, land, verify that the landing spot can easily be dug into, dig a hole, install the sensor, and finally, fly away.
Originally, the Nebraska Intelligent MoBile Unmanned Systems (NIMBUS) Lab presented a paper at IROS that outlined a quadcopter with the ability to carry an auger with an embedded sensor and plant it into the ground. A few weeks later at ISER, they presented another paper on how the drone can autonomously determine whether the landing spot is a good digging spot.
The biggest challenge to designing this drone was keeping it in the air despite having a drilling rig, sensors, and computers, which is all necessary, yet heavy, equipment for the drone to carry around. This is especially important since the whole point of the drone is to place sensors in locations that are particularly hard for humans to navigate.
In response, NIMBUS Lab came up with a solution: a helicarrier, a parachute, and a unique drone deployment system. The project, known as UAS Digging and In-Ground Sensor Emplacement, has provided a video of the interesting concept.
A large UAS carries a smaller UAS in a capsule to permit long-range transport, according to Carrick Detweiler, co-director at NIMBUS Lab in an interview with IEEE Spectrum. Once the drones reach a spot near the goal location, the smaller UAS drops and parachutes down to the ground.
“We use a parachute as this allows the capsule to be dropped from most any altitude. When it is low enough, the capsule releases, then the arms on the UAS extend before detaching from the parachute to fly to the goal location,” said Detweiler, according to IEEE. “We travel with the arms folded up to reduce the size of the capsule to make the whole system more aerodynamic. The [smaller] UAS is based on a heavily modified DJI S1000 using with a Pixhawk flight controller to allow autonomy. Our UAS then flies to the goal location to dig a hole in the ground to emplace a sensor.”
Detweiler said they chose this deployment system since battery-powered drones have shorter flight times. In order to address this problem and navigate to distant locations, the smaller UAS needed to hitch a ride somehow. In the video, they used a large, gas-powered UAS with extended hours of flight time, but a manned aircraft or other systems could also carry the smaller UAS.
A potential use for these autonomous drones are navigating to places such as wetlands that are hard to get to and hard to plant sensors in mushy ground.
“Our system can be used to deploy a wide range of sensors in remote or hard to access locations. We have a USDA-NIFA project where we are deploying sensors and UASs in sensitive wetlands environments, which are often hard to access in other ways without impacting the environment,” said Detweiler. “We need to dig the sensors into the ground both to secure the sensors so they don’t get washed away, but also for sensors such as soil moisture sensors that need good contact below the surface.”
Additionally, the systems aren’t perfected yet. Detweiler pointed out that unpredicted weather made the lab consider some adjustments for the future of these drones.
“Field experiments with real systems are always challenging. In this case it was particularly challenging as our system was carried over 10 miles and in weather we had not previously tested in (high winds and below freezing temperatures),” said Detweiler. “This emphasized the need for greater autonomy systems that can automatically monitor to detect faults and take corrective actions in situations that the programmers may not have initially envisioned. We will be continuing to work on this system to increase the robustness to uncertain conditions.
The researchers also believe beyond sensor placement, the drones could be used to team up for range extension and payload delivery in areas such as medical supply delivery and delivering more drones.
Filed Under: Rapid prototyping
