RoBoat V1

The RoBoat project is done in conjunction with Rassweiler lab at FSU. The main goal of the project is to create an autonomous surface vessel that can be used as a platform for data collection devices. With secondary goals of being low cost and easily portable. Version 1 consists of the initial design and local field tests up until the first field usage in Mo’orea in May 2021

 

 

Hardware

All designs for the 3D printed parts can be found on Thingiverse.

Frame

Initial Wooden Frame

Upgraded Extruded Aluminum Frame

The primary design concerns around the frame were that it needed to be small enough to be portable, break down for shipping and large enough to support the 3 camera configuration needed for generating photo mosaics. With those in mind the initial frame was designed to be 36×38 inches with the pontoons being 55 inches. To test the load capacity an initial prototype was built with wood. Tests were done where weights were added to measure max load of the craft. After initial tests proved the design would work 3030 extruded aluminum was used to replace the wood frame. For the pontoons 4 inch irrigation PVC was used with 3d printed nose cones created to provide aero dynamics. The frame is attached to the pontoons using 3d printed mounts that are secured to the pontoon using hose clamps. The use of  extruded aluminum and hose clamps allows for easy disassembly and the overall size is wide enough for the camera rig but small enough to fit fully assembled in a vehicle.

 

Nose Cone
Pontoon Mount

Motor Assembly

For the propulsion of the craft, the two design considerations were minimizing the draft and making sure the motors survive salt water. This is because the craft will be primarily used around salt water coral reefs. Since the craft would be used in salt water Blue Robotics T100 motors were chosen for propulsion due to their affordability and reliability in salt water. To mount the motors to the pontoons a custom motor mount was 3d printed to connect to the T100 and then that mount connects to a pontoon adapter. The two piece design was chosen so that way if a different motor was used with different mount points only the motor mount would need to be replaced. The pontoon adapter connects to the pontoon using 2 hose clamps. To protect the motors from impact and fins were created and placed directly in front of the motors.

 

Pontoon Adapter
Motor Mount

Fin

Camera Assembly

To generate the photo mosaics a GoPro needs to be mounted to each side of the craft and in the middle. When the craft is in water the cameras need to go deep enough that they do not go above water during wavy conditions but also be able to be retracted when the craft is on land or if the craft runs aground. To achieve this a swing arm mount was created that can use a gravity lock to keep the cameras up when on land but can easily be lowered when the cameras need to be in the water. The swing arm design allows the cameras to automatically retract back in case of impact to protect the cameras.

 

Side camera Mounts
Center Camera Mounts

Electronics

Electronics Connection Diagram

For autonomous control the boat, Pixhawk 2.4.8 was chosen using Ardupilot Rover software. To allow for onboard programing, switch control and to act as a wifi access point a Raspberry Pi Zero was connected to the pixhawk. Four 6000mAh 4s Lipo batteries were used in parallel as a power source. 

 
 
 

Power Switch

To provide a hard cutoff of power, a RV power switch was used that will physically disconnect power from the batteries to the electronics. As part of the power switch assembly, 2 water proof switches were added that connected to the Raspberry Pi. The switches served as a load waypoints button and a start mission button.

 

Power Switch Enclosure

 

 

Software

All source code for this project can be found on Github. The Raspberry Pi communicates to the Pixhawk using Mavlink protocol.

 

 

Results

Video 1: Local Gulf trials before trip to Mo’orea

Video 2: First Mo’orea test run post assembly

Video 3: Going over previous years path

Video 4: GoPro footage of Benthic Track

 

 

 

Improvements

After several missions the motors sucked in seaweed and since the autopilot was trying to continue its course it kept trying to increase power to the motors eventually causing the motor to overheat. This not only caused damage to the motor but also damaged the motor connector to the electronics case. Further study revealed that T100 thrusters are only rated at max power for short bursts and that Blue Robotics offers a newer thruster the T200 that can continually hold max power. So switching to the T200 should prevent motor damage if seaweed clogs the motor. The other improvement to stop this issue will be to manufacture guards that go over both the front and back of the motor to prevent seaweed from getting in the motor in the first place.

Another issue that was encountered was that the waterproof switches started to fail after a few uses in the saltwater. The work around was to disconnect them and just use the wifi hotspot to connect to the boat from a phone using the QGroundcontrol app to start the mission.