Narrative:

I was working satellite combined; VFR weather. A BE40 entered my airspace via fails; putting him on a modified right base for the ILS 24 at cgf. An EA50 arrived from a 'straight in' position. I don't recall exactly how much faster the BE40 was compared to the EA50; but it was enough to warrant; in my mind; giving the EA50 a vector to the south to build room to put the BE40 in front of him. I accomplished this relatively easily; and I remember thinking that I had been a bit conservative in waiting as long as I did to put the EA50 on the localizer. They were about four miles apart and the BE40 was showing a higher ground speed than the EA50; who was trailing. For the sake of ensuring that my 4 mile interval would remain; I instructed the BE40 to maintain 180 KTS or greater until krook; the IAF. The EA50 was instructed to maintain 180 KTS or less. I don't recall exactly what ground speed the BE40 showed as he crossed krook; I believe it was around 210 KTS. In any case; within a few seconds his displayed ground speed had dropped to about 160-170. The EA50 was showing an overtake; I think around 30 KTS. I was surprised that a scenario which I had seen countless times before seemed to be developing differently this time. Both aircraft were talking to the cgf tower; and I mistakenly believed that they could provide visual separation between the two aircraft. I had forgotten that I was required to call cgf and request that they provide visual separation. In any case; I'm not sure I could have effected the coordination before separation was lost. Cle recently changed to fusion; replacing; or rather; supplementing; our ASR-9. One of the stranger things myself and others have noticed is that it often takes the software a bit of time to accurately display ground speed. A side effect of this is that we often see impossibly large decreases in ground speed. I have observed aircraft showing a ground speed in excess of 400 KTS; only to watch it decrease drastically to less than 300 KTS in a matter of seconds. I believe I felt a false sense of security as I observed the BE40's ground speed; thinking he was much faster than he actually was and possibly waiting too long to assign a speed because he seemed to be going fast enough on his own. I've since tried to be more aware of discrepancies between displayed ground speed; and my expectations based on instructions; winds; and aircraft performance. And of course I brushed up on the LOA with cgf and the application of visual separation. I do feel; however; that the speed issue with fusion is significant. Cle uses only one long range radar; and one ASR-9; plus the ads-B to generate the fused targets. There seems to be a lot of 'guessing' by the software; resulting in strange phenomena such as ptls pointing in different directions than the aircraft's apparent line of travel; targets that appear to jump around or wander slightly; and impossibly large decreases in ground speed display. One can tell the difference between three or four 'guess' hits by the fusion software; followed by one 'real' hit from the ASR-9. I think awareness of these issues will go a long way to helping controllers work around them; but my guess is that until ads-B becomes much more prevalent; or cle gets another radar feed or two to supplement the two we have now; our fused targets may be less reliable than most. I'm not an expert on the subject; I'm just reporting what I've observed.

Related reports:

Google
 

Original NASA ASRS Text

Title: CLE Controller described a loss of separation event claiming the new 'Fusion' RADAR software sometimes displays inaccurate ground speeds.

Narrative: I was working satellite combined; VFR weather. A BE40 entered my airspace via FAILS; putting him on a modified right base for the ILS 24 at CGF. An EA50 arrived from a 'straight in' position. I don't recall exactly how much faster the BE40 was compared to the EA50; but it was enough to warrant; in my mind; giving the EA50 a vector to the south to build room to put the BE40 in front of him. I accomplished this relatively easily; and I remember thinking that I had been a bit conservative in waiting as long as I did to put the EA50 on the localizer. They were about four miles apart and the BE40 was showing a higher ground speed than the EA50; who was trailing. For the sake of ensuring that my 4 mile interval would remain; I instructed the BE40 to maintain 180 KTS or greater until KROOK; the IAF. The EA50 was instructed to maintain 180 KTS or less. I don't recall exactly what ground speed the BE40 showed as he crossed KROOK; I believe it was around 210 KTS. In any case; within a few seconds his displayed ground speed had dropped to about 160-170. The EA50 was showing an overtake; I think around 30 KTS. I was surprised that a scenario which I had seen countless times before seemed to be developing differently this time. Both aircraft were talking to the CGF Tower; and I mistakenly believed that they could provide visual separation between the two aircraft. I had forgotten that I was required to call CGF and request that they provide visual separation. In any case; I'm not sure I could have effected the coordination before separation was lost. CLE recently changed to FUSION; replacing; or rather; supplementing; our ASR-9. One of the stranger things myself and others have noticed is that it often takes the software a bit of time to accurately display ground speed. A side effect of this is that we often see impossibly large decreases in ground speed. I have observed aircraft showing a ground speed in excess of 400 KTS; only to watch it decrease drastically to less than 300 KTS in a matter of seconds. I believe I felt a false sense of security as I observed the BE40's ground speed; thinking he was much faster than he actually was and possibly waiting too long to assign a speed because he seemed to be going fast enough on his own. I've since tried to be more aware of discrepancies between displayed ground speed; and my expectations based on instructions; winds; and aircraft performance. And of course I brushed up on the LOA with CGF and the application of visual separation. I do feel; however; that the speed issue with FUSION is significant. CLE uses only one long range RADAR; and one ASR-9; plus the ADS-B to generate the fused targets. There seems to be a lot of 'guessing' by the software; resulting in strange phenomena such as PTLs pointing in different directions than the aircraft's apparent line of travel; targets that appear to jump around or wander slightly; and impossibly large decreases in ground speed display. One can tell the difference between three or four 'guess' hits by the FUSION software; followed by one 'real' hit from the ASR-9. I think awareness of these issues will go a long way to helping controllers work around them; but my guess is that until ADS-B becomes much more prevalent; or CLE gets another RADAR feed or two to supplement the two we have now; our FUSED targets may be less reliable than most. I'm not an expert on the subject; I'm just reporting what I've observed.

Data retrieved from NASA's ASRS site as of July 2013 and automatically converted to unabbreviated mixed upper/lowercase text. This report is for informational purposes with no guarantee of accuracy. See NASA's ASRS site for official report.