4 ms·

This "research" has a glaring statistical error, principally that the researchers assumed that UAV locations are either uniformly distributed or follow the same

by maxcan 11y ago

This "research" has a glaring statistical error, principally that the researchers assumed that UAV locations are either uniformly distributed or follow the same distributions as birds. The problem, especially for small aircraft, is that the sites which attract concentrations of small aircraft - like Golden Gate Bridge, Hudson River, etc, are the same places which attract UAV operators. Also, they seem to neglect the growth in populations of UAV. UAVs are exploding in popularity and capabilities. These numbers are going to get worse and the FAA is taking its usual caution-first approach.

ArkyBeagle 11y ago
They are very very thorough and cautious. That's probably what we want, ultimately.
Houshalter 11y ago
I wonder what the true risk of flying a drone is. Lets say their number is 50 times greater than realistic. The 1.87 million years divided by 50 would be a more reasonable 37,400 years of flight time per accident. Assuming every plane has 200 passengers, that's about, or about 6.105×10^-7 deaths per hour of flight time. If I did my math right, that means it's about 0.6 micromorts per hour. By comparison, the average person has about 1.6 micromorts of unnatural causes of death per day, so it is a fairly large relative increase in risk. If you were to require every drone operator to pay for insurance, assuming an economic value of life at about $9.1 million, then it would cost $5.56 per hour of flight time. Not cheap!
- matheweis 11y ago
  As with many things, there are a number of factors. Just of few of these include the weight and speed of the drone, the location that it is being flown, the competency of the operator, and so on... You're also assuming that a drone vs aircraft incident would result in the complete loss of the aircraft. I would imagine that we will eventually see such an incident, but it probably won't result in catastrophic loss...
  - Houshalter 11y ago
    Well we are doing a (very rough) estimate of the average expected risk of flying drones. Obviously some people would be far more riskier than others. As for the number, I just took that from the article. It's probably an overestimate of the risk, and certainly most planes don't have 200 passengers.
Animats 11y ago
Right. I noticed that. They're assuming that drones are distributed randomly across the entire volume of US airspace. The analysis isn't based on data from pilot reports. The FAA tallies drone sightings from aircraft, and for 2015, they had 650 reports through August 9th.[1] That's a lot, considering that seeing a small drone from an aircraft means it is rather close. Drone reports are available online.[2] Most of the reports are from near airports. Here are a few months of reports from LAX alone: PRELIM INFO FROM FAA OPS: LOS ANGELES, CA/UAS INCIDENT/091658P/LOS ANGELES ARRIVAL REPORTED SOUTHWEST 366, B737, REPORTED SEEING AN UNIDENTIFIED HELICOPTER AT 4,000 FEET WHILE ON A 6 MILE FINAL TO RUNWAY 24R. LOS ANGELES PD AIR SUPPORT ADVISED, LOS ANGELES PD 18 RESPONDED. PRELIM INFO FROM FAA OPS: LOS ANGELES, CA /UAS INCIDENT/1821P/SOUTHERN CALIFORNIA TRACON REPORTED SOUTHWEST 366, B737, OBSERVED A QUADCOPTER TYPE UAS AT 4,000 FEET 6 MILE FINAL LAX. LAPD AVIATION UNIT NOTIFIED. PRELIM INFO FROM FAA OPS: LAX/UAS INCIDENT/0815P/LAX ATCT ADVISED COMPASS 5680, E170, SEA-LAX, REPORTED UAS 200 FEET ABOVE ACFT ALTITUDE OF 600 FEET OFF RIGHT WING WHILE ON 2.5 MILE FINAL RWY 24R LAX. LA CO SHERIFF NOTIFIED. "PRELIM INFO FROM FAA OPS: LAX/UAS INCIDENT/1625P CESSNA C525 OBSERVED AN REDDISH-ORANGE-YELLOW COLORED UAS, MODEL UNKN, WHILE SOUTHBOUND AT FL200. NO EVASIVE ACTION TAKEN. "PRELIM INFO FROM FAA OPS: LOS ANGELES, CA/UAS INCIDENT/1357P/LAX ATCT ADVISED UNITED 701, B757, PAX REPORTED TO PILOT SEEING A UAS TURN TOWARDS ACFT ON LEFT SIDE WHILE ON FINAL RUNWAY 24R, 2,000 FEET. Summary: POSSIBLE DRONE SIGHTING: UAL701 (B757), LANDED RY24R & LATER CALLED LAXT ON AN UNRECORDED LINE TO REPORT WHAT 2 PASSENGERS ADVISED HIM: ON 6-MILE FINAL TO RY24R, THE PASSENGERS CLAIMED TO SEE A DRONE BTWN 2000' - 2100' BTWN RY24L & RY24R TURN TOWARD THE AIRCRAFT; NO INJURIES REPORTED. "PRELIM INFO FROM FAA OPS: LOS ANGELES, CA/UAS INCIDENT/2017P/UNITED 1600, B737, REPORTED SEEING A UAS FLY ABOVE ACFT AT 2,600 FEET 8 E LAX. LAPD AVIATION AIR SUPPORT WAS NOTIFIED AT 213-485-2600. NO EVASIVE ACTION TAKEN. Summary: DRONE REPORT, UAL935, B772, ON FINAL RY25L. UAL935 REPORTED A DRONE OFF THEIR LEFT SIDE ON APPROXIMATELY A 1 MILE FINAL AT 300-400FT. PILOT, CAPT. JEFF GLEASON, RPTS WHITE QUAD DRONE (ABOUT THE SIZE OF A MEDIUM PIZZA BOX), STATIONARY JUST WEST OF THE 405 PASSED JUST UNDER THE LEFT WING. NO EVASIVE ACTION TAKEN. DEN NTFD. PRELIM INFO FROM FAA OPS: LOS ANGELES, CA/UAS INCIDENT/0851P/LAX ATCT ADVISED JET BLUE 23, A320, REPORTED A WHITE UAS ON THE LEFT SIDE OF THE ACFT WHILE WSW BOUND AT 500 FEET 1 E LAX. NO EVASIVE ACTION TAKEN. LAPD AERO BUREAU NOTIFIED. Summary: DRONE REPORT FROM JBU23 (A321) 1 MILES EAST OF LAX, LEFT (SOUTH) OF RY25L FINAL @ 500 FEET, STATIONARY WHITE DRONE, NO EVASIVE ACTION, LA COUNTY SHERIFF AERO BEREAU NOTIFIED @ 1610Z, DEN NOTIFIED @ 1611Z. That's just LAX. The distribution of reports is nowhere near random. In addition to airports, drones have been reported near stadiums; one crashed into a spectator. The paper is from George Mason University's Mercatus Center, which is a right-wing "university source for market-oriented ideas." The authors are economics students. It's not from anybody involved in aviation safety. [1] https://www.faa.gov/news/updates/?newsId=83445 https://www.faa.gov/news/updates/?newsId=83445 [2] https://www.faa.gov/news/updates/?newsId=83544 https://www.faa.gov/news/updates/?newsId=83544

by maxcan 11y ago

ArkyBeagle 11y ago

They are very very thorough and cautious. That's probably what we want, ultimately.

Houshalter 11y ago

I wonder what the true risk of flying a drone is. Lets say their number is 50 times greater than realistic. The 1.87 million years divided by 50 would be a more reasonable 37,400 years of flight time per accident. Assuming every plane has 200 passengers, that's about, or about 6.105×10^-7 deaths per hour of flight time. If I did my math right, that means it's about 0.6 micromorts per hour. By comparison, the average person has about 1.6 micromorts of unnatural causes of death per day, so it is a fairly large relative increase in risk. If you were to require every drone operator to pay for insurance, assuming an economic value of life at about $9.1 million, then it would cost $5.56 per hour of flight time. Not cheap!

matheweis 11y ago
As with many things, there are a number of factors. Just of few of these include the weight and speed of the drone, the location that it is being flown, the competency of the operator, and so on... You're also assuming that a drone vs aircraft incident would result in the complete loss of the aircraft. I would imagine that we will eventually see such an incident, but it probably won't result in catastrophic loss...
- Houshalter 11y ago
  Well we are doing a (very rough) estimate of the average expected risk of flying drones. Obviously some people would be far more riskier than others. As for the number, I just took that from the article. It's probably an overestimate of the risk, and certainly most planes don't have 200 passengers.

Animats 11y ago

Right. I noticed that. They're assuming that drones are distributed randomly across the entire volume of US airspace. The analysis isn't based on data from pilot reports. The FAA tallies drone sightings from aircraft, and for 2015, they had 650 reports through August 9th.[1] That's a lot, considering that seeing a small drone from an aircraft means it is rather close. Drone reports are available online.[2] Most of the reports are from near airports. Here are a few months of reports from LAX alone: PRELIM INFO FROM FAA OPS: LOS ANGELES, CA/UAS INCIDENT/091658P/LOS ANGELES ARRIVAL REPORTED SOUTHWEST 366, B737, REPORTED SEEING AN UNIDENTIFIED HELICOPTER AT 4,000 FEET WHILE ON A 6 MILE FINAL TO RUNWAY 24R. LOS ANGELES PD AIR SUPPORT ADVISED, LOS ANGELES PD 18 RESPONDED. PRELIM INFO FROM FAA OPS: LOS ANGELES, CA /UAS INCIDENT/1821P/SOUTHERN CALIFORNIA TRACON REPORTED SOUTHWEST 366, B737, OBSERVED A QUADCOPTER TYPE UAS AT 4,000 FEET 6 MILE FINAL LAX. LAPD AVIATION UNIT NOTIFIED. PRELIM INFO FROM FAA OPS: LAX/UAS INCIDENT/0815P/LAX ATCT ADVISED COMPASS 5680, E170, SEA-LAX, REPORTED UAS 200 FEET ABOVE ACFT ALTITUDE OF 600 FEET OFF RIGHT WING WHILE ON 2.5 MILE FINAL RWY 24R LAX. LA CO SHERIFF NOTIFIED. "PRELIM INFO FROM FAA OPS: LAX/UAS INCIDENT/1625P CESSNA C525 OBSERVED AN REDDISH-ORANGE-YELLOW COLORED UAS, MODEL UNKN, WHILE SOUTHBOUND AT FL200. NO EVASIVE ACTION TAKEN. "PRELIM INFO FROM FAA OPS: LOS ANGELES, CA/UAS INCIDENT/1357P/LAX ATCT ADVISED UNITED 701, B757, PAX REPORTED TO PILOT SEEING A UAS TURN TOWARDS ACFT ON LEFT SIDE WHILE ON FINAL RUNWAY 24R, 2,000 FEET. Summary: POSSIBLE DRONE SIGHTING: UAL701 (B757), LANDED RY24R & LATER CALLED LAXT ON AN UNRECORDED LINE TO REPORT WHAT 2 PASSENGERS ADVISED HIM: ON 6-MILE FINAL TO RY24R, THE PASSENGERS CLAIMED TO SEE A DRONE BTWN 2000' - 2100' BTWN RY24L & RY24R TURN TOWARD THE AIRCRAFT; NO INJURIES REPORTED. "PRELIM INFO FROM FAA OPS: LOS ANGELES, CA/UAS INCIDENT/2017P/UNITED 1600, B737, REPORTED SEEING A UAS FLY ABOVE ACFT AT 2,600 FEET 8 E LAX. LAPD AVIATION AIR SUPPORT WAS NOTIFIED AT 213-485-2600. NO EVASIVE ACTION TAKEN. Summary: DRONE REPORT, UAL935, B772, ON FINAL RY25L. UAL935 REPORTED A DRONE OFF THEIR LEFT SIDE ON APPROXIMATELY A 1 MILE FINAL AT 300-400FT. PILOT, CAPT. JEFF GLEASON, RPTS WHITE QUAD DRONE (ABOUT THE SIZE OF A MEDIUM PIZZA BOX), STATIONARY JUST WEST OF THE 405 PASSED JUST UNDER THE LEFT WING. NO EVASIVE ACTION TAKEN. DEN NTFD. PRELIM INFO FROM FAA OPS: LOS ANGELES, CA/UAS INCIDENT/0851P/LAX ATCT ADVISED JET BLUE 23, A320, REPORTED A WHITE UAS ON THE LEFT SIDE OF THE ACFT WHILE WSW BOUND AT 500 FEET 1 E LAX. NO EVASIVE ACTION TAKEN. LAPD AERO BUREAU NOTIFIED. Summary: DRONE REPORT FROM JBU23 (A321) 1 MILES EAST OF LAX, LEFT (SOUTH) OF RY25L FINAL @ 500 FEET, STATIONARY WHITE DRONE, NO EVASIVE ACTION, LA COUNTY SHERIFF AERO BEREAU NOTIFIED @ 1610Z, DEN NOTIFIED @ 1611Z. That's just LAX. The distribution of reports is nowhere near random. In addition to airports, drones have been reported near stadiums; one crashed into a spectator. The paper is from George Mason University's Mercatus Center, which is a right-wing "university source for market-oriented ideas." The authors are economics students. It's not from anybody involved in aviation safety. [1] https://www.faa.gov/news/updates/?newsId=83445 https://www.faa.gov/news/updates/?newsId=83445 [2] https://www.faa.gov/news/updates/?newsId=83544 https://www.faa.gov/news/updates/?newsId=83544