Why am I excited about these boats?

I'm excited about these boats for a number of reasons! They are low-cost and open-source: so any interested person or NGO can purchase the parts, and put them together at an affordable cost! They are solar-powered and completely autonomous: so once they are launched they can be left alone! They require minimal maintenance: sensors will need to be recalibrated only yearly! They have a full complement of available sensors to monitor water-health: the only thing lacking are sensors to monitor bacteria in the water. They can provide real-time data corresponding to water-health, transmitted via cellular data networks, which can subsequently be shared on websites!

I would like to see these boats deployed by local people in their nearby lakes and reservoirs: perhaps to provide evidence that a suspicious factory really is altering the physicochemical properties of the water! The boats, for instance, could be programmed to 'hang out' near the factory outlets/runoffs! Another task I would like to see the boats undertake is a complete sampling of a river from source to sea: this might not be a task for which real-time monitoring via cellular networks is appropriate, but the boats can easily be adapted to log data and dock at assigned ports to physically off-load memory cards.

At present the robot pollution-detection boat has three systems:

1. Autonomous navigation and collision-detection system
2. Water-health sensor platform
3. Propulsion and hull

1. - Autonomous Navigation and Collision-Detection System

HARDWARE

This employs the arduino uno microcontroller, with a motor shield, and a GPS shield. For collision-detection an ultrasonic distance sensor is installed at the front of the boat. The motor shield drives a servo to control the rudder, and a motor to power the propeller

SOFTWARE

Outline of autopilot navigation

1. Parse GPS’s output NMEA sentences with GPS.parse(GPS.lastNMEA()) to obtain GPS.latitude, GPS.longitude and GPS.angle
2. Set current latitude and longitude e.g. currentlat = GPS.latitude currentlon = GPS.longitude
3. Set the waypoint we want to reach next e.g. destlat for destination latitude, destlon for destination longitude
4. Calculate distance from currentlat and currentlon to the destlat and destlon using haversine formula
5. Calculate bearing/heading from currentlon to destlon (heading)
6. Compare bearing/heading that was calculated with current bearing from gps (i.e. GPS.angle = currentheadingfromgps)
7. currentheadingfromgps - heading value will inform us of how to turn the rudder e.g if going straight we can do myservo.write(90)
8. Keep turning the rudder until currentheadingfromgps comes within 1 degree of heading, then stop turning
9. Keep propeller going until distance to waypoint is <=0.5 metres, then stop propeller
10. Remain at waypoint for desired length of time (1hr?) then repeat

Next we might want to employ the PID (proportional-integral-derivative) control method to keep the boat on a more accurate path to the waypoint; although I am dubious if it's really necessary for this project!

2. -Water-health sensor platform

This sensor platform which utilizes the arduino mega will contain sensors for water health, a GSM shield to transmit the data from sensors to a PC and ultimately the internet, and a solar panel for power

DISSOLVED OXYGEN (DO) SENSOR

-Why monitor DO?

Most polluted rivers suffer from low levels of dissolved oxygen. Dissolved oxygen in a river can vary from 0 to 18mg/L. It’s generally accepted that 4-5mg/L are required to support a healthy population of fish; anything below 3mg/L will not be enough to ensure survival of fish, and levels lower than 4mg/L will negatively impact fish reproduction. All species have their own particular DO requirments: worms and fly larvae and algae can for instance tolerate low DO, whilst larger fish may require > 6mg/L

Factors which affect levels of dissolved-oxygen

1. Water temperature: higher temperatures will ‘push’ out the oxygen due to increased movement of water molecules [Note: polluted rivers...