Floating robotic devices to monitor river pollution

China is reported to be using drones to detect polluting factories. Good idea! http://timesofindia.indiatimes.com/Home/Environment/Pollution/China-uses-drones-to-detect-polluting-factories/articleshow/47332399.cms

A related development could be using robotic “boats” to monitor river pollution. It is very tempting to call them “roboats”! Factory effluents and sewage enter rivers at specific points, and it needs constant policing to detect and deter criminal pollution of rivers. Roboats could simplify the work.

The challenge is in ensuring a very high chance that the roboat could travel a required stretch of the river without getting stuck in weeds or rocks or getting grounded on the shore somewhere. We need to look for some design that ensures that the roboat would prefer to stay in that part of the river stream where the velocity is highest.  Is it true that a smooth spherical object, perhaps weighted to reduce tumbling, would avoid the slow moving side of the water due to the Venturi effect? It is worth checking this out. The roboat could be a smooth sphere made of a suitable material, ideally one that will let cell phone signals travel through it. We will need sensors in contact with water to make suitable measurements. A transmitter based on a smart phone could be inside the shell and send out signals whenever it is able to sense the presence of signals from a base station. We could also build in mechanisms that could help the roboat to wriggle free if it gets stuck.

An alternative design would be to use an aluminium alloy shell and mount a water proof control and communication device on top. A suitable weight inside the shell could provide an orientation to the device to keep the concealed antenna on top. GPS sensing could be used to provide location information and data might need to be stored till connectivity is available for transmission. It might be useful to recover the roboat before the river falls into the sea, but we should also be willing to treat it as a dispensable object. The data it would provide during a run, for instance along a stretch of River Ganges, would be worth a lot more than the cost of a device like that. Developing inexpensive sensors to make appropriate measurements would be a key part of the project. It would be great to create enough electrical energy from the physical motions of the device to keep the control/communications device charged. Another idea worth exploring is that of a wireless beacon on the device that gets turned on at the end of the run, to facilitate its recovery. 

We could also use tethered devices to monitor specific locations, including locations off the sea coastline.