Recent progress in drone technology has opened up a wide range of applications to cost-effective drones that be flown by non-professionals. Advances in object avoidance, battery technology and sensor miniaturization enable powerful drones that can be flown by pretty much anyone over distances of miles, measuring and recording various observations. Coupled with Artificial Intelligence (AI), these drones can be extremely helpful in many different scenarios. These scenarios can range from construction to disaster relief and saving one’s life. In my case, I will be using drones for environmental applications to hopefully prevent natural disasters.
A key for any environmental application is the ability to monitor, measure and record various parameters at their source. Whether it's air pollution measurements of PM2.5 particles, water pollution monitoring due to oil spills, or smoke/flame monitoring for wildfire detection, real-time sensor measurements at high-risk locations are key. As such, a mobile sensor that can “visit” the high-risk location and perform and record measurements can be very invaluable. As such, I have been working on attaching sensors to drones and having the drone communicate wirelessly in real-time to my previously architected wireless sensor network. Such an architecture would enable the sensor network to become mobile and perform real-time measurements at various high-risk locations. AI algorithms in the background can process this sensor data and try to “predict” when something might go wrong.
Although I am using the drone for three different aspects, prevention is the most important. If prevention is completed successfully, there will be no worry of any disaster, and the problem can be solved before it even starts. For example, a drone could come across a corroded powerline. By leveraging Machine Learning and pre-trained AI models, the drone would identify the risk of corrosion and the possibility of creating sparks. Or it could notify that the transformer at that exact location needed to be fixed and state how quickly the problem would need to be fixed.
Another important factor is early detection. For example, this can be conducted around oil rigs by waterproof drones. These drones can guide themselves along pipes and lines and look for cuts or gashes that lead to significant leaks. This constant drone vision would detect a tear and quickly report it to local authorities. Multiple drones would be put to work, increasing the chance of early detection and nullifying and form of late detection.
In conclusion, drones can be beneficial in many different scenarios and for various reasons. I will be using a drone to prevent and detect natural disasters such as oil spills and wildfires. The drones can fly along with high voltage equipment and swim right next to pipes to check for sparks, cuts, ruptures, and more. The drones would be trained with a significant amount of data and predict when high-voltage equipment would need maintenance. They would also be fed data to identify cuts or gashes in the pipes that require immediate assistance. The drones would be deployed in large numbers and almost eliminate any slowness of detection or prevention.
Comments