Robotics Institute 

University of Michigan, Ann Arbor

© 2017 by Bruce Personal Website. 

Updated on 8/21/2017

 

Outline & Link

 

Multirotor

A multicopter is a mechanically simple aerial vehicle whose motion is controlled by speeding or slowing multiple downward thrusting motor/propeller units. The most popular multirotor on the market is the quadcopter, which is fast, easy to manufacture, and affordably priced. A quadcopter features four propellers arranged symmetrically with vertically oriented motor shafts. Moreover, multicopters do not require an airstrip. However, quadcopters are not powerful enough to deliver heavy loads. Furthermore, quadcopters are not well-suited to carry valuable items, as they are grounded if just one of the four motors fails.

 

A hexacopter is the next step up from a quadcopter. Hexacopters have six motors and corresponding propellers. This adds to the aircraft’s capabilities and really makes for a more optimal choice for anyone seeking to attach expensive cameras or convey heavy loads. Essentially, these models have all of the benefits of a quadcopter, while equipped with even more powerful capabilities.

 

  • Power: Hexacopters have higher speeds and more power due to the two extra motors.

 

  • Flying altitude: Hexacopters are able to fly higher with relative ease.

 

  • Safety: Hexacopters have six motors positioned 120 degrees apart. Should one motor fail, the others will simply pick up the slack. This means that a pilot will be able to safely land the machine even if one motor is damaged, or fly it to a repair station, instead of immediately being grounded.

 

For my project, I have decided to create a hexacopter with a wheelbase of 810mm. The large size allows for a heavier load and more stable flight to better capture high-quality images, render 3D models. The extended battery life derived from automated wireless charging allows the hexacopter to carry high-performance CPUs and GPUs as well as two cameras, one of which is a depth camera that can create a 3D map of a forest robotically.

 

Let us consider the following scenario. A person orders something over the Internet. Information on both the product and the buyer are sent to the multicopter via Wifi. However, the product is stored too far away for the multicopter to arrive without needing to charge its battery. The hexacopter of my project would use GPS to locate a charging pad and recharge wirelessly, thus expanding its range. A multicopter outfitted with a standard battery can fly for about 30 minutes at 25 miles per hour. If a charge is permitted, we can extend the range from 12.5 miles one way to 60 miles or even 80 miles in one day by charging five times and flying six times, making 75 miles feasible within one day, assuming a four-hour charging period.

 

The GPS information, detailing the best route from starting point to charging pad, can utilize the flying control system Pixhawk. Generally, there is a roughly 50 cm error for a GPS sensor. Therefore, I have put an omni-antenna on the charging pad to ensure its location is known. I have also placed a high-directivity antenna with a motor and an Arduino computer on the hexacopter, ensuring that the high-directivity antenna rotates. This is similar to how ships at sea find a lighthouse. Moreover, I have used IR track to reduce calculations and enable the hexacopter to land on the wireless charging pad precisely. Moreover, the hexacopter can calculate its power consumption from the start point to the charging pad. This means the hexacopter will be able to calculate how much it needs to be charged to be able to travel to the next charging pad. If it determines that it has been sufficiently charged to reach the next pad, it departs. If not yet sufficiently charged, it remains at its position for further charging.

 

The major problem of using multicopters—limited distance due to battery factors—is thereby ameliorated. Other applications can also be developed. For example, I have used a high-resolution camera, GoPro Hero4, to record images, and the hexacopter could also be made to create 3D map models with Matlab or Computer Vision and use depth cameras. This could help ascertain the best locations for more charging pads.