Drones are getting larger and autonomous
Airports don’t have capacity
Even the skies won’t have capacity
We normally think of a drone as a small machine carrying at most a few kg of payload. And traveling a short distance, at most a few kms.
All that has changed in the last one year. And the rate of change is exploding. Take this announcement:
It will soon be commonplace for drones to be large. They may carry people, host hospitals or transport cargo. Once they get electrified, they will be cheap to use. A classic example is the real photo of a drone hospital for disaster mgmt. This cabin could be fitted in a bus (real bus photo enclosed) or just be capable of flying around on its own. With machines and people inside it.
This is going to create havoc in the skies. You don’t want these drones colliding or falling apart. What happens to them if you have fog or bad weather. Autonomous drones will steer off course. The answer is not to build weather proof drones because there are no such possibilities.
The answer lies in better weather monitoring and prediction. Jaishri Sanwal Bhatt introduced me to Prof Srinivasan at the Jawaharlal Nehru Institute of advanced studies. He has been using ai and ML to predict fog and bad weather conditions at Bangalore airport. And his ai model is being trained. Last week his model was at 75% accuracy. The challenge is getting accurate time series data for a prolonged period of time. If they could get it for the entire planet they could extrapolate the model to not just detect bad landing conditions but also to prevent lightning strikes or turbulent weather
I reached out to Kumar Margasahayam who recommended we use data from the Spire constellation of two hundred nano cube satellites. Any spot on earth is covered every 24 minutes. And time series data is available for 30 years. If this is integrated as part of the training data for the weather ai platform of the IIT Alumni Council to complement ground work done in fog prone Delhi Airport and monsoon prone Bangalore airport – drone and flying objects can be made a lot more safer.
Experts seem to opine that the Air India Ahmedabad crash was driven by a battery burnout which cut off the fuel. However, a significant number of other aircraft accidents seem to have a weather issue.
Real time images from satellites can solve this. It is just a question of connecting the dots.
But someone has to do it.
IIT Alumni Council seems best placed
Finally drones will not just be autonomous, they will need neuromorphic computing to course correct when thrown off course. This is where inertial navigation systems come in. The gyroscopes will drift and gps sensors have to correct this drift. The tejas fighter jet uses an indigenous inertial navigation system from Ashvani Shukla – another iit alumnus. So autonomous aircraft will require this live feed. Portals like weather.com are not doing this as yet.