ROV Quard Patrol System

There is a large fish farm on our lake. It's made from a 900-foot-long fishing net and a height of 30 feet. The fish grows to 40 pounds and often tears the net. Our diving club is located on the shore of this lake and we have to patrol the fishing net several times a day and check it for damage. We decided to make an automatic robot for this tedious work. Look how simple it is to make an automatic underwater Arduino based patrol system with WiFi controller.

Automatic patrol along the specified path
Continuous video recording, programmable photography

Payload - up to 4 pounds

Operating mode - continuous

WiFi programming all modes and transferring data and video directly from under the water

Speed is from 0.1 to 10 miles per hour

Automatic speed correction depending on the flow of water

Speed Correction by Markers

Programming of the motion algorithm: continuously, run / pause, hourly, daily and any other combinations

Automatic transmission of footage

The maximum depth is 600 feet

Battery charge - automatic (underwater charger with inductor)

Does not require additional maintenance for a long time

We have already done all this and have started testing :)

1. a piece of PVC pipe 80mm od - $12 (more length means more buoyancy, we used 1 feet)

2. 3x18650 batteries - $25 (Panasonic 3400 mAh is the best for now)

3. battery protection/charge board - $10 (very usful)

4. Arduino mega - $10~$30 (original or copy, mega or uno)

5. WiFi module - $8

6. DT700 brushless motor - $14

7. DIY ROV thruster 3D printed nozzle for DT700 motor DIY ROV kayak surf 4 kg truster - $40 (or at Amazon)

8. Inductive charger - $10

9. ESC brushless motor controller with reverse - $14

Let's see how it all works.

Our robot guard goes along a long fishing line. On the edges of the line are attached magnets and the robot has two Hall sensors. Thus, the robot moves from the magnet to the magnet. Algorithm and speed of movement is programmed through Arduino.

In order for our robot to be recharged directly under water we used two large inductors. One is installed at the top of the robot, the second is on the underwater base. They work well even in salt water and give a charge current of 0.8-0.4 Ampere at a distance of 1-2 inches.

The hull made of a thick-walled PVC pipe so that our robot could withstand a lot of pressure. The edges of the pipe are covered with lids with o-rings. Inside the pipe there are 3x18650 batteries, Arduino with a WiFi module and a motor speed controller. The wires for the motor, Hall sensors and inductor are fused through the pipe wall. The robot body also has additional outputs for powering and controlling the external flashlight and camera.

The WIFI antenna is connected to the metal guide rings. When the robot arrives at the base, the ring touches the antenna wire, which rises to the surface. Thanks to this, we can control the robot directly under water.

We used one of the best thruster. At just 0.5 Ampere it gives incredible thrust of 1.1 pound. That means it can go at 3 miles per hour with a big underwater camera for 6 hours without recharging. Or, for example, it can range 660 feet in 3 minutes, and then fully recharge in 5 minutes.

For the test, we took a line of 120 feet and a minimum speed. The robot drove there and back for one minute and recharged every hour, from time to time we established contact with him on the WF and changed the program. During the day there was not a single failure and he made more than 500 trips! The next phase of testing will already be on our fishing farm. Our lake will soon be completely covered with ice and the inspection of fishing nets would be difficult. We are going to stretch a long line along the fishing net. The near edge of the line will be at the very surface of the water. To begin with, our robot will run along the network once per hour, each time returning to the base. If necessary, we can easily lift it to the surface for maintenance and repair. If everything works properly, the robot will spend the whole winter under water and make several thousand flights automatically. Later I hope to publish a report about:)