Artificial Intelligence - A.I.

Table of Contents

Drone A.I.

The first question that comes to mind is why do we need AI in drones? The simple answer to this would be to make them smarter. Depending on your use of drones, you might want then to be capable of far more complex tasks than taking off, navigating via remote control or landing.

Drones are now playing a key problem-solving role in a variety of sectors, including defense, farming, natural disaster relief, security, and production. With their ability to increase productivity and improve safety, drones have become important tools for everyone from the military to farmers.

Until recently, though, drones were only able to display what their cameras captured. Now, thanks to artificial intelligence software, they can perceive their surroundings, which enables them to map areas, track objects and provide analytical feedback in real-time Thanks to AI software, drones can now treat what they see and communicate back in real-time. This new generation of drones, for instance, can map up to 2.7 million square miles (an area roughly as large as the contiguous 48 U.S. states). Additionally, the military deploys them in war zones and emergency response teams such as firefighters fighting forest or wildfires or use them in containment and recovery operations.

In order to have effective drones, we need to be able to reduce human intervention in the data collection process. We need to make intelligent drones that can read, calculate, analyze and predict data themselves to provide useful information. Without Humans, drones can rely on the in-built machine learning algorithms to operate.

Identify: Core to any AI algorithm to fully function would be its ability to identify and match what it sees. To recognize a person from a tree or a dog from a car. Once it is able to distinguish objects and/or people, we then have the foundation for more advanced processes. Take for example the simple follow me functions most commercial drones have. Before ir can perform this command it first has to be able to identify you or its controller. Once this is done, it can then move to the next stage which, in some cases, might involve maintaining a fixed distance and altitude from you for a set period of time.

Calculate: Here the AI is tasked with calculating certain variables in the environment as relating to objects that it has identified. This might involve its distance from said object, elevation, inclination, number count and other relevant data stream.

Analyze: Once all this raw data has been compiled it need to break it down into what it can discard, and what it needs to further process based on its mission objective and corresponding environment. Take for example a farming drone. If it is tasked to spray insecticide on a given plot of land at certain intervals, it would have to analyze the size of the plot and make out which has crops growing on it and which doesn’t so as not to incur waste.

Predict: If we take one step further with our previous example we will find that it is possible to have an advanced AI that not only knows which part of a plot to spray but also which one not to based on previous data regardless if there are crops within that given section. Another case might be a set of military drone that is tasked with following a given convoy if they predict that there is no chance of the convoy stopping might decide to spread themselves to cover every possible path that the convoy might take instead of a single path. Drones can process sensor data and plot its way ahead by analyzing the obstructions in its’ way. One of the popular Machine learning algorithms that can be used for this purpose is known as ‘Fuzzy logic’.

These simple concepts would turn a fun toy used by kids the world over into an efficient and sort after tools that can cover a wide variety of uses and scenarios. Do not be fooled into thinking that the limitations of use for Drone AI fall primarily to the Military and security institutions. As stated above both farming, construction, medical and logistics sectors have just as much to gain with the adaptation of the technology.


2. Reena Kamra. (2019). AI-powered Drones: A technological benison. Available: Last accessed 28/10/2019.
3. Malek Murison. (2019). Anyone Can Design a Hybrid Drone With MIT CSAIL’s AI Platform. Available: Last accessed 28/10/2019.

Visit next page to learn about - Robotics