Sense and Avoid: Analysis of Sensor Design Factors for Optimal Deconfliction

Transcription

1 Sense and Avoid: Analysis of Sensor Design Factors for Optimal Deconfliction Basically, we want this: For these:

2 Background: UAVs Weight Mid-Sized UAVs The Big Ones Small UAVs MAVs The area of study for this Analytical Report Speed-Range-Endurance

3 Sense And Avoid How can we sense an impending collision?!!! We need to exchange information, either visually, electronic telemetry, noise, etc. When we have information about each other, we can negotiate our paths. However, a pilot in an aircraft cannot readily be apart of this exchange: 1) A larger, manned aircraft has the right of way over the UAV anyway. 2) A pilot already has to process a lot of information, and: 3) Even if we warned the pilot about a collision threat, they couldn't see the UAV easily, or be able to react with the same reflexes. Therefore, the collicion-avoidance decision and maneuver are one-sided.

4 !!! SAA Methods Electromagnetic Radiation Passive Radio VHF Voice Transponder Radar (2nd-ary) Active Radar (Primary) T-hertz Possible area, not treated Visible Regular Image Reflected Laser Spectral Signature Acoustic Information Passive: Engine Noise Flight Plans Airways Jet Routes Active: Sonar Particle Radiation Alpha, Beta, Neutron, etc Possible, but not treated X-Rays Not treated Gamma

5 Analysis: Electromagnetic Medium Type Qualitative Attributes Determination VHF Voice Radio 108 Intermittent: Only used by pilot near other air-traffic, airports. Not reliable: Antenna too big? Transponder (Passive) 1090 Intermittent: Requires aircraft to be interrogated by TCAS or ATC Radar. Not reliable: must have nearby ATC Active Transponder 1090 Possible: We can cause any nearby aircraft to transmit a return. Formidable design process, investigate Radar (Secondary) 1030 Intermittent: Requires aircraft to be illuminated by surveillance radar. Not reliable Not plausible Radar (Primary) Any UAV carrys an on-board radar. Not plausible Regular Image nm Using cameras, we can see the image of a offending aircraft. Requires on-board computing. Requires intensive computing power, many variables Spectral Signature nm If we know the spectral signature of a typical offending aircraft, we can tune an appropriate imager to see Same as regular image; sensors are very expensive F or λ

6 Analysis: Acoustic Medium Type F or λ Engine Noise 20 Hz Most air-traffic makes noise. Traffic 10kHz speed determines report-to-conflict time-span. If below feet, traffic under 250 knots (~290mph) Best Option, Investigate Further Sonar Any Not reliable: must have nearby ATC Qualitative Attributes Determination Difficult: Requires UAV to transmit a sound pulse, time for report return doubles.

7 Analysis: Information Information Type Qualitative Attributes Determination Air-traffic Flight Plan Only if air-traffic is IFR (Instrument Flight Rules), basically, jets. Subject to change, human-handled. Too vague for use by UAV collision avoidance, impractical. Jet-ways, Traffic Routes Near airports and other areas, airtraffic is confined into narrow corridors. Defined on air charts. Useful for general collision risk mitigation. Not useful in impending collision.

8 Conclusions Active Transponder 1090 Possible: We can cause any nearby aircraft to transmit a return. Engine Noise 20 Hz Most air-traffic makes noise. Traffic 10kHz speed determines report-to-conflict time-span. If below feet, traffic under 250 knots (~290mph) Formidable design process, investigate Best Option, Investigate Further Why is this useful? Allow small UAVs to have access to National Airspace: Possibilities for airborne chemical sensors (Find a natural gas leak before it explodes, discover meth labs, etc) Cinematography Others Make Aviation Safer: Cheap sensors for UAVs means slightly more expensive but effective manned-aircraft sensors.

9 Drat! Already Done!