How would you design an airport traffic control system.

Question

Anonymous · Accepted Answer

Assuming we are talking here about the experience of flight arrivals/departure
from a specific airport.  Few key data points to be considered while designing
this experience:
a. Weather of the specific area
b. Availability of runway to land or take-off
c. proper functioning of the communication devices between pilot and ATC
d. In air assistance to the pilots
e. Pilot to be able to smoothly report air/weather conditions back to a central
system from where then it can be relayed to other flights passing through the
same area.
While all other use cases are seem to be handled, e is the one where I feel
there is a lot more that can be worked on referencing the SIA incident happened
recently.
 Let's design a central system where pilots should be able to send real time
weather climatic feeds. It could be a server with high computing capabilities as
the data from this server will then need to be relayed to multiple other flights
passing through the area. One can think of this server as a self sufficient
system with all automation processes running inside it. The way it will work is:
a. Capture all the feeds from the flights in air
b. it should have access to data of all the flights of specific areas, ofcourse
there will be mutiple such system deployed for load balancing, DR, security and
other privacy reasons
c. All these servers should be able to communicate with each other as well
d. Any critical (Red/yellow) information to be relayed to the flights passing
specific Area.
e. This should be a next-gen system which should pro-actively warns pilot/flight
if any specific area has seen critical incident happening or likely hood of Air
incidents when passing through the area.
f. The system should be able to help pilot navigate through turbulance if
already caught like what height, speed and angle to Manov our the flight to come
out of the turbulance.
g. Should be able to send SOS signals to the nearby airports/flights if the
system has identified that the flight will need emergency landing assistance.

Designing such system will require extensive Infrastructure as well as thorough
testing and should be piloted first with small amount of flights in the region
with less air traffic and then a slow roll out per region should be an ideal way
with catching and fixing bugs along the way.

Success Metrics:
North star: Accurate predictions
Number of times it relayed critical weather alert to the pilot also capturing
false positives
Number of times it triggered SOS, capturing false positives
How many airports have adopted this
With scale are we seeing drop in accuracy
How many flights a single server can successfully monitor
Time it takes to process information and then send it back to pilots
Availability of the system itself

Pricing:
Given the monster system to be designed, this is likely to be pricier as the
infrastructure it will need and software that need to scale with flights, people
travelling across the board etc. At a very high estimate a single server may
cost from 100K to 300K

How would you design an airport traffic control system.

Did you come across this question in an interview?

Answers

Try Our AI Interviewer