Search and Rescue Drone

Purpose.

When natural disasters strike remote areas, getting aid to those affected can be difficult. Searching for survivors is often done by helicopter, but this is costly and slow. Therefore, knowing exactly what areas need resources the most can greatly improve the quality of response.

Our team designed an autonomous drone platform that could identify the location of survivors, estimate their conditions, and relay information quickly to first responders. These drones could operate individually or in swarms in a variety of inclement weather conditions.

There are already drones marketed as search-and-rescue tools, but these are expensive and have limited flight times. We designed our drone to cost less than $500 and to have a flight time of at least one hour.

 Use Scenarios.

To better understand the problem, our team created use cases detailing the drone’s operation in specific scenarios. Here are three of them.

• Flooded Area Response:

• Night Time Search Response:

• Lost Hiker/Camper Response:

Benchmarking.

Researching what solutions already exist gave a us a good sense of what design combinations would yield success. It also helped us determine what engineering requirements we should prioritize.

These plots show the relationship between select attributes for a range of drones already on the market. We needed our drone to have a long flight time, which would require a high battery capacity. A battery with a high capacity would not only cost more but also weigh more, making it harder to achieve long flight times. From the plots below, it is observable that a flight time of 60 minutes is doable with a battery capacity of less than 4000mAh, which is within our price range. According to the plots, this is only achievable with a fixed wing design.

Preliminary Design.

A fixed wing design can sustain the desired flight time, but only once it’s in the air. For our drone to be operable in remote locations, it will need to take off and land where there are not clear runways. This is, admittedly, where rotary wing aircraft (e.g. helicopters) excel, which is why they are often used for search and rescue. So, our drone would ideally take off and land vertically but transition to horizontal fixed wing flight.

Working off of group brainstorms, I used Procreate to sketch concepts of our drones. Below are the three finalist designs.

Design Revision.

We settled on a fixed-wing configuration with a rotor set into each wing for VTOL capabilities. However, preliminary weight loading analyses indicated that the configuration would be heavy and require huge batteries.

Inspired by the WingtraOne surveying drone (below), we pivoted our design to a tailsitter configuration. This way, the drone could take off vertically, but wouldn’t require any specialized subsystems.

If our drone was to be a viable tailsitter, it would need the right airfoil. I led the airfoil selection process, where the S7055 (below) was chosen from a list of possibilities. Mathematical analysis showed that this airfoil would facilitate long flight times but would not obstruct VTOL operations.

In order to simplify manufacturing, we wanted to have as few control surfaces as possible. We opted for a semi-elliptical wings with large elevons. The drone would have slow maneuverability, but the trade-off of fewer moving parts and less weight was deemed desirable.

Using publicly-available airfoil property data, combined with the preliminary planform design, analyses were conducted to determine lift force versus drag force in a variety of conditions. Good engineers design for worst case scenario, so in the analysis the drone speed was set to 35km/h, which was 10km/h less than the desired cruising speed. At least mathematically, the wing design yielded a very high lift force-to-drag force ratio, as is visible below.

To ensure the airfoil and planform met our needs, an airflow analysis was performed using SimScale. The cross-section in the leftmost image below displays the airspeed distribution around the wing root, the rightmost image displays the same conditions for the tip.

Below are the two finalists for our tailsitter configuration, drawn in Procreate. The final design used the configuration seen on the left, but included the rangefinders pictured in the right one in the tips of wings.

CAD Modeling.

With a design configuration finalized, the other mechanical engineer, Mutahir, and I modeled the design to scale using SolidWorks. I rendered the final model with emergency livery using Keyshot.

Fabrication.

The ribs that gave the wing its characteristic airfoil shape were lasercut from Baltic birch, which was chosen due to its light weight and high strength. Spanwise spars and the elevon axles were made from carbon fiber. The wing design was modified to accommodate and extra spanwise spar, lending stiffness to the wing without adding much additional weight. Mounts for the propeller motor and elevon servo were 3D printed from PLA.

Takeaways.

Unfortunately, the project didn’t yield a finished product. This was partly because of the COVID-19 pandemic, which hit halfway through our first semester.

But it was also because the project was overly ambitious: there was little chance that a team of six people with no prior experience could design an aircraft - even a small one - from the ground up. However, because of this, we learned much more than we ever would’ve imagined. We had to stretch ourselves into unfamiliar areas of knowledge, and work together as a team to understand concepts that we hadn’t been taught.

Obviously, with a client, you don’t have the option of not finishing. However, I know now that good engineers and designers are willing to stray far outside of their fields just to bring back the knowledge needed to do a project right.

Want To Learn More?

This page only scratches the surface of our capstone project. For an even more in-depth look into the details and processes of the Autonomous Search and Rescue Drone, use this QR code to check our project’s wiki page here!