4 ms·

Synthesizing position from multiple inputs sounds like a process analogous to a Kalman filter.

• eximius 11y ago
  ... Or common sense. You know, don't put all your eggs in one basket, etc.
  • theoh 11y ago
    Actually I thought it might add value to the discussion by making the comparison. We're mostly computer engineers here, after all. But never mind.
    • dalke 11y ago
      It's very little if anything like a Kalman filter. Or rather, as much like a Kalman filter as it is a PID controller. For example, Nainoa Thompson described part of learning to be a navigator from Mau Piailug. The full account is at http://pvs.kcc.hawaii.edu/ike/hookele/locating_land.html http://pvs.kcc.hawaii.edu/ike/hookele/locating_land.html . In short: > The diurnal flights of such birds are the most useful signs for expanding landfall, since their flights to and from an island gives a fairly specific direction to the wayfinder. As the birds leave an island in the morning, the wayfinder can sail in the direction the birds are coming from to find land; as the birds return to an island in the late afternoon, the wayfinder can follow the birds to land. In this case: > The night before that we saw the birds flying south so how come late in the morning with the sun very high was this bird coming out of the north? That would suggest that we passed the island during the night. But this interpretation was wrong: > Mau had seen in the beak of the bird a little fish and he knew that the birds were nesting. They had flown out to sea before sunrise and were taking food back at mid-morning to feed their young, before they flew out to sea again to feed themselves. How is this meaningfully like a Kalman filter?