Using ADSB Data to track performance.

[Shadrach]

Does anyone have any insight as to what degree free ADS-B Data can be trusted for performance analysis? 

On a recent flight we elected to takeoff and do a cruise climb (~110mph) to 13,500MSL.   I thought I'd take the opportunity to analyze real world climb performance as we're planning a Coast to Coast trip next summer. 

Both flightaware and ADS-B showed that we climbed from a pressure altitude of 1100' to 12,950' for an 11,850' gain in 12mins and 46 secs.  That works out to 928fpm 752fpm.  I am trying to derive some real data in which I can be confident. DA was a few hundred feet below standard. We had about 600lbs of people and fuel (~450 under gross).   I am not confident in the data.  What am I missing in regards to the limitations of using these tools to derive performance real world performance data?

 

EDIT: Turns out the data was not the problem. My spreadsheet entries were. New number is a still surprising but more believable 752fpm.

[Ragsf15e]

I think you can use them pretty effectively if you’re realistic in data collection.  The biggest factors affecting climb are weight and power.  The aircraft is pretty sensitive to weight and since adsb doesn’t have that info, it’s all on you.  For power, there are myriad things that could affect it, but the temp is going to make a significant difference, especially the temps aloft.  Everything about the engine and propeller is less efficient with that high density altitude, so you’ll have to normalize that somewhat.  Finally, turbulence is going to affect your climb and speed somewhat, so only count smooth data.

 I believe there’s an faa or aopa app for doing this.  Flight performance monitor or ga performance or something like that.

[Shadrach]

5 minutes ago, Ragsf15e said:

I think you can use them pretty effectively if you’re realistic in data collection.  The biggest factors affecting climb are weight and power.  The aircraft is pretty sensitive to weight and since adsb doesn’t have that info, it’s all on you.  For power, there are myriad things that could affect it, but the temp is going to make a significant difference, especially the temps aloft.  Everything about the engine and propeller is less efficient with that high density altitude, so you’ll have to normalize that somewhat.  Finally, turbulence is going to affect your climb and speed somewhat, so only count smooth data.

 I believe there’s an faa or aopa app for doing this.  Flight performance monitor or ga performance or something like that.

It was a butter smooth day even under the broken deck, so that and perhaps some gentle updrafts off the western ridges may have contributed to what I thought was an overly ambitious number for the temp and pressure.

[Ragsf15e]

2 minutes ago, Shadrach said:

It was a butter smooth day even under the broken deck, so that and perhaps some gentle updrafts off the western ridges may have contributed to what I thought was an overly ambitious number for the temp and pressure.

Check this, i bet you’ll do better on your own, but the faa is trying…

https://www.faa.gov/news/safety_briefing/2018/media/SE_Topic_18-12.pdf

i tried this a while back and didn’t love the app.

Your climb is definitely better than I would have expected in my F.  You were light so that helped, but I would have expected 1100 fpm for the first few thousand feet, then decreasing to 500-700 fpm by 10,000.

[Shadrach]

25 minutes ago, Ragsf15e said:

Check this, i bet you’ll do better on your own, but the faa is trying…

https://www.faa.gov/news/safety_briefing/2018/media/SE_Topic_18-12.pdf

i tried this a while back and didn’t love the app.

Your climb is definitely better than I would have expected in my F.  You were light so that helped, but I would have expected 1100 fpm for the first few thousand feet, then decreasing to 500-700 fpm by 10,000.

Thanks for the link.

I was cruise climbing into the wind. It was not a max climb performance effort which gave me all the more reason to question the number.

[EricJ]

ADS-B out, which feeds the data you're using, generally uses WAAS, so is pretty accurate.    Since you're using long-term analysis (e.g., large altitude changes over reasonably long time) the small errors get averaged out better.   So I think it's very usable as you're using it.    The usual caveats of interpreting ground speed and wind effects doesn't matter as much for the climb performance.

[Fly Boomer]

On 4/28/2023 at 11:33 AM, Shadrach said:

Does anyone have any insight as to what degree free ADS-B Data can be trusted for performance analysis? 

On a recent flight we elected to takeoff and do a cruise climb (~110mph) to 13,500MSL.   I thought I'd take the opportunity to analyze real world climb performance as we're planning a Coast to Coast trip next summer. 

Both flightaware and ADS-B showed that we climbed from a pressure altitude of 1100' to 12,950' for an 11,850' gain in 12mins and 46 secs.  That works out to 928fpm.  I am trying to derive some real data in which I can be confident. DA was a few hundred feet below standard. We had about 600lbs of people and fuel (~450 under gross).   I am not confident in the data.  What am I missing in regards to the limitations of using these tools to derive performance real world performance data?

The only thing I would add is to be cautious about moving the cursor back and forth on that graph.  You are likely getting a single ADS-B data point.  I would maybe look at the detailed list of ADS-B data points, and average 10 or so consecutive points to help smooth out the data.  Also, I suspect that some of those data on FlightAware are interpolated from surrounding data.

[Shadrach]

54 minutes ago, Fly Boomer said:

The only thing I would add is to be cautious about moving the cursor back and forth on that graph.  You are likely getting a single ADS-B data point.  I would maybe look at the detailed list of ADS-B data points, and average 10 or so consecutive points to help smooth out the data.  Also, I suspect that some of those data on FlightAware are interpolated from surrounding data.

So, I found an error in my spread sheet.
 

The new calculation is a much more reasonable 752fpm. 11,850 in 946 secs.
 

I only posted the graph because an image of the raw data with time signatures was too large to capture in a practical way. I used the two data points below to calculate time to climb.

Time (EDT)             Latitude     Longitude     Course   kts    feet       Rate

Wed 10:48:05AM  39.6979      77.7303        ↓188°     77    1,100     935 

Wed  11:03:51AM    39.6792     77.9094        → 83°    138  12,950   566  

 

 

[Fly Boomer]

1 minute ago, Shadrach said:

So, I found an error in my spread sheet.
 

The new calculation is a much more reasonable 752fpm. 11,850 in 946 secs.
 

I only posted the graph because an image of the raw data with time signatures was too large to capture in a practical way. I used the two data points below to calculate time to climb.

Time (EDT)             Latitude     Longitude     Course   kts    feet       Rate

Wed 10:48:05AM  39.6979      77.7303        ↓188°     77    1,100     935 

Wed  11:03:51AM    39.6792     77.9094        → 83°    138  12,950   566  

 

 

Still pretty impressive.  Might be informative to calculate rate of climb in two- or three-thousand-foot blocks to show the down low climb vs the "up there" climb.

[Shadrach]

8 minutes ago, Fly Boomer said:

Still pretty impressive.  Might be informative to calculate rate of climb in two- or three-thousand-foot blocks to show the down low climb vs the "up there" climb.

Might be. However lower altitude climb was a leisurely cruise climb so not very telling.

MAPAs Evaluation numbers to 10,000 were 702fpm single pilot with full tanks, so similar weights but different conditions.

I need to do a few max performance time to climb over different conditions but at the same weight to derive meaningful data.

[A64Pilot]

I don’t understand, what’s wrong with ships altimeter and clock?

[Shadrach]

18 minutes ago, A64Pilot said:

I don’t understand, what’s wrong with ships altimeter and clock?

Not a thing.  However if there is very detailed data available at no charge, why not utilize it?
I was skeptical of my original calculation. Turns out, the problem was me and my failure to properly transfer the ADSB data to a spread sheet.

[Hank]

15 minutes ago, A64Pilot said:

I don’t understand, what’s wrong with ships altimeter and clock?

No glitz, no glamor . . . . But that’s how I've done it. Best done with someone to write as I call out numbers, especially right after takeoff.

That's another benefit to the modern way--you can do it at your desk a day or two after the flight, and concentrate on the numbers and not on flying the plane, checking MP, VSI, ASI and Altimeter and writing them all down correctly. 

[wombat]

@Shadrach check out my spreadsheet that uses both ForeFlight and ADS-B data to calculate fuel economy.  It could do climbs too, but I wasn't really paying attention to that at that time.

 

https://docs.google.com/spreadsheets/d/1ixfZQPAsKIh59g-zvnmV05KYbNdMHMJ3dWYyBAGOaMM/edit?usp=sharing

[Fly Boomer]

4 minutes ago, wombat said:

@Shadrach check out my spreadsheet that uses both ForeFlight and ADS-B data to calculate fuel economy.  It could do climbs too, but I wasn't really paying attention to that at that time.

 

https://docs.google.com/spreadsheets/d/1ixfZQPAsKIh59g-zvnmV05KYbNdMHMJ3dWYyBAGOaMM/edit?usp=sharing

I have not looked at your doc yet.  How do you get the data in?

[wombat]

On FlightAware, you can look at the track log and it has the exact time and position data.     ForeFlight does as well, so you can just cut-n-paste them into a single spreadsheet and insert empty cells as necessary for the times to line up.

If I set the time correctly in my engine monitor, that would be even easier. But it's off by a little so I had to shift it manually too

Oh, and this data is for a brand C aircraft.    It's just to demonstrate that the analysis can be done

Edited May 1, 2023 by wombat

[Ragsf15e]

9 hours ago, Shadrach said:

So, I found an error in my spread sheet.
 

The new calculation is a much more reasonable 752fpm. 11,850 in 946 secs.
 

I only posted the graph because an image of the raw data with time signatures was too large to capture in a practical way. I used the two data points below to calculate time to climb.

Time (EDT)             Latitude     Longitude     Course   kts    feet       Rate

Wed 10:48:05AM  39.6979      77.7303        ↓188°     77    1,100     935 

Wed  11:03:51AM    39.6792     77.9094        → 83°    138  12,950   566  

 

 

Sounds reasonable.  Was it maybe pretty cool out too?  These numbers seem much closer to what I get in my F.

[Shadrach]

11 minutes ago, Ragsf15e said:

Sounds reasonable.  Was it maybe pretty cool out too?  These numbers seem much closer to what I get in my F.

Close to standard, DA was 300’ below field elevation when we departed.