3 ms·

ok, lets talk stats. Lets NOT assume he is 99.99% safe. Lets look at the 825,000 drones sold in one year, and estimate it from the number of problems that the

by myrryr 8y ago

ok, lets talk stats. Lets NOT assume he is 99.99% safe. Lets look at the 825,000 drones sold in one year, and estimate it from the number of problems that they had with those. We don't just pick a number when we have a perfectly good sample to work with. I AM looking at the population at large. If there are There are 2.5 million drones in operation at the moment, which I am happy to agree to, and we are not seeing many problems, how much of a sample do you need?

madrox 8y ago
99.99% is a wild assumption merely to illustrate a fallacy. If you want better probabilities, I have none, since so much is dependent on external factors. Applying a blanket probability to "how safe is it to fly a drone" is meaningless. But almost 2000 drone incidents were reported in 2017 alone. For airplanes in that same time, unless I'm wildly misreading the FAA website, is...7. That's isn't quite apples to apples since the criteria for inclusion in each as an incident is different, but this is HN and not an FAA study. Point still being that even considering how unlikely something seems you need to compare it to the base rate.
- jfoutz 8y ago
  After reading this thread, i thought of a sort of obvious solution. Require some amount of liability insurance. There are all sorts of complex regulations that could be put in place, but this is an easy, low effort one.
  - sokoloff 8y ago
    Are you assuming that the insurance industry will sufficiently regulate safety to achieve a safe overall outcome? (Note that this is in significant parts true today in the manned, piston-engine aircraft world.) Or just suggesting insurance so that the injured get paid?
    - jfoutz 8y ago
      I’m suggesting it’s a very lightweight regulation. We are in this weird time, where actual safety is hard to ascertain. The safest solution is to simply ban drones. But, there might be some utility there. Building up a history of risk would be super helpful. Some things are obvious. A six ounce toy is less dangerous than a 12 rotor monster. But that toy can still cause problems when it’s floating over the freeway. I think insurance would create a data collection environment that could drive more realistic, useful regulations. FWIW, movie studios generally require 5 million liability for drone shots. It’s pretty cheap to purchase, but those pilots are professional. Expanding to amateurs is complicated. Insurance gives some visibility into that complexity.
  - cmurf 8y ago
    There is no regulation requiring private pilots obtain insurance. Such a requirement comes from the bank whose loan made it possible to acquire said plane; and most any airport with hangars will require evidence of it. I'm not sure what buying insurance does. It certainly provides no standardization of rules or training. Have you read any FARs? There are FARs for flying kites and ultralights. What about any of that do you think is so complicated? Feel free to use a regulation in FAR 91 as an example for something onerous that insurance can obviate. One of my favorite regulations in these UAV conversations is: § 91.119 Minimum safe altitudes: General. Except when necessary for takeoff or landing, no person may operate an aircraft below the following altitudes: (a)Anywhere. An altitude allowing, if a power unit fails, an emergency landing without undue hazard to persons or property on the surface. There is no reason at all why a drone should ever hurt someone, other than sheer incompetency and lack of imagination. Now where are you going to get that imagination? And how are you going to obtain competency? Insurance doesn't get you either of those things.

by myrryr 8y ago

madrox 8y ago

99.99% is a wild assumption merely to illustrate a fallacy. If you want better probabilities, I have none, since so much is dependent on external factors. Applying a blanket probability to "how safe is it to fly a drone" is meaningless. But almost 2000 drone incidents were reported in 2017 alone. For airplanes in that same time, unless I'm wildly misreading the FAA website, is...7. That's isn't quite apples to apples since the criteria for inclusion in each as an incident is different, but this is HN and not an FAA study. Point still being that even considering how unlikely something seems you need to compare it to the base rate.

jfoutz 8y ago
After reading this thread, i thought of a sort of obvious solution. Require some amount of liability insurance. There are all sorts of complex regulations that could be put in place, but this is an easy, low effort one.
- sokoloff 8y ago
  Are you assuming that the insurance industry will sufficiently regulate safety to achieve a safe overall outcome? (Note that this is in significant parts true today in the manned, piston-engine aircraft world.) Or just suggesting insurance so that the injured get paid?
  - jfoutz 8y ago
    I’m suggesting it’s a very lightweight regulation. We are in this weird time, where actual safety is hard to ascertain. The safest solution is to simply ban drones. But, there might be some utility there. Building up a history of risk would be super helpful. Some things are obvious. A six ounce toy is less dangerous than a 12 rotor monster. But that toy can still cause problems when it’s floating over the freeway. I think insurance would create a data collection environment that could drive more realistic, useful regulations. FWIW, movie studios generally require 5 million liability for drone shots. It’s pretty cheap to purchase, but those pilots are professional. Expanding to amateurs is complicated. Insurance gives some visibility into that complexity.
- cmurf 8y ago
  There is no regulation requiring private pilots obtain insurance. Such a requirement comes from the bank whose loan made it possible to acquire said plane; and most any airport with hangars will require evidence of it. I'm not sure what buying insurance does. It certainly provides no standardization of rules or training. Have you read any FARs? There are FARs for flying kites and ultralights. What about any of that do you think is so complicated? Feel free to use a regulation in FAR 91 as an example for something onerous that insurance can obviate. One of my favorite regulations in these UAV conversations is: § 91.119 Minimum safe altitudes: General. Except when necessary for takeoff or landing, no person may operate an aircraft below the following altitudes: (a)Anywhere. An altitude allowing, if a power unit fails, an emergency landing without undue hazard to persons or property on the surface. There is no reason at all why a drone should ever hurt someone, other than sheer incompetency and lack of imagination. Now where are you going to get that imagination? And how are you going to obtain competency? Insurance doesn't get you either of those things.