Bayes Rule used to find lost Malaysian Plane
Malaysian Airline Flight 370 (not to be confused with Malaysia airlines flight 17, which was lost this summer) disappeared March 8th 2014 flying from Kuala Lumpur to Beijing. 239 people were aboard when it lost contact about an hour into the flight. The last known location of the flight is an area over the Gulf of Thailand.
Bayes Rule has been used to help triangulate the location of the downed flight. In fact this equation has been used to find many lost planes including Air France Flight 447, which crashed in 2009. The search for Air France’s black box “triggered the most expensive and exhaustive search effort conducted for a plan1“. Two years after the plane had crashed the area in which the plane may have crashed was only narrowed down to an area the size of Switzerland. Then Bayes Rule was recruited to help with this problem, and just five days later the black box was found.
Bayes Rule has proved to be useful in these situations because the equation is dynamic and continuously updates with new information about the location of a lost plane. Using this, probabilities are then calculated about a certain area, and based on that probability the area is then searched (or not searched). This theorem has many real world applications. Bayes Rule is used by Google’s driverless cars, it makes predictions in the stock market and it even helped to locate German U-boats during World War II.
Some information that was used to calculate these probabilities includes wind currents, water currents, previous flight patterns and underwater drift. In this case Bayes Rules seemed to not be of much help. The plane, after eight months of searching, is still lost. The French plane was likely found sooner, after including Bayes Rule, because of the huge amount of information that had already been collected and was available.
Links:
http://en.wikipedia.org/wiki/Malaysia_Airlines_Flight_370