What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

RV accident data analysis

jmbaute

Well Known Member
I wanted to learn a new data analysis tool (Power BI) so I decided to build a simple dashboard based on the NTSB database. I thought it might be interesting to the community so I am sharing the link with you.

s!AhLtUuPGOPDSvr8kxsktth8ISBHoXQ

edit: supposed to be a pic here but for the life of me I can't figure out why it isn't showing.

This is the same data you'll find in lots of other places, including on this website (Doug's page was actually my inspiration).

A couple notes:
- the data source is the XML file on the NTSB site. They update the data once per day. At this time I'm not sure if the link I'm posting refreshes against that data or if I have to trigger that refresh on my side.
- I'm using some custom logic against the entire database of ~81k incidents to get to "Vans" models. This is a little tricky since as you know, you can call your aircraft anything you want at registration. I'm "close" to what Doug reports on his page and for now I'm calling it "good enough".
- The report is fairly interactive, for example click on the model donut to filter the heat map and grid. Same thing on the heat map, click to filter by phase of flight. Click again to clear your filter.
- If you click any of the links to the final report and don't get a result, it is because there isn't one issued yet. I'm just building that URL based on the accident number.
- I'm certainly open to feedback/suggestions but this is just something I'm playing with so I can't promise I'll do anything else with it.
- I don't care if you share the link.
- If a moderator wishes to move or delete this post, I'm fine with that, too.

Blues Skies,
John
 
Last edited:
Just visited your page: this is outstanding and should be a great tool for anyone who wants to study RV accidents.
 
Nice work!

I agree.......thanks for doing this...
But this is something I will not be showing my wife! :eek:
 
All GA accidents

Here is a similar report showing the same data for the GA fleet at large, for the data nerds out there (amateur built not included in this).

Doug, thanks for posting the screenshot on the front page of VAF, if there is value, feel free to add the interactive chart to your accidents page; code snip below, although we are getting slightly different numbers. I'd be happy to discuss what filters I'm using.

Code:
<iframe width="800" height="600" src="https://app.powerbi.com/view?r=eyJrIjoiMTI1OTk1YmItZGQ1ZS00Mzg2LTk4MjMtOWIyNDFjMWVlYWQxIiwidCI6ImU0MmUzNTJmLWZiZDUtNGI3Mi1iOTRlLTFjMDkzNzk0Mjc4MiIsImMiOjN9" frameborder="0" allowFullScreen="true"></iframe>
 
Does this data take into account flight hours or fleet sizes? If not, what value does this data provide?
 
Flight Hours

Does this data take into account flight hours or fleet sizes? If not, what value does this data provide?
I agree - that was the first thing that popped into my mind - how many hours flown? Unfortunately that's a really hard number to get it seems. It can be estimated, but there are no firm numbers.
 
Does this data take into account flight hours or fleet sizes? If not, what value does this data provide?
You're right; there are no denominators, which is what makes the difference between data and anecdotes.

Nonetheless, I think it has tremendous value as a collection of anecdotes. For instance, having read something like 50 reports on ground loops, I've learned that it isn't generally the initial deviation that causes the accident, but the overcorrection. Most of them read like, "Upon landing, the airplane encountered a gust and veered to the left. The pilot corrected with right rudder and brake, causing the airplane to veer to the right. He attempted correction with left rudder and brake but was unable to regain control. The airplane departed the runway to the right....."

I'm not trying to start a conversation about groundloops, just illustrating that these reports are valuable, even if statistically meaningless.
 
I agree - that was the first thing that popped into my mind - how many hours flown? Unfortunately that's a really hard number to get it seems. It can be estimated, but there are no firm numbers.

I read accident stuff for two reasons - one is to get an idea of the RATE of accidents, and in that case, this data doesn?t;t tell us anything.

But the more important reason for reading accident stuff is figuring out the CAUSE of accidents - and in that case, this is really fascinating information. Knowing what causes accidents can help us avoid them ourselves. Knowing how many (or how often) accidents happen isn?t really useful, except to bean counters.

TRhis is pretty neat stuff - but as I always say, you really need to read the descriptions of accidents to help learn to avoid them yourself.
 
Does this data take into account flight hours or fleet sizes? If not, what value does this data provide?
No, because they aren't reported in the NTSB data set the report uses.

Maybe "analysis" isn't the right word to describe this effort, because the analysis is really up to the end user. It is more of an easy way to search/slice the individual incidents.
 
John B - funny that you should create this I started doing it in excel based on a post a few days ago. What might be fun is to see the "A" vs tail wheel comparisons and also the % of incidents per model as a percentage of models flying from Vans website. The 6's have been around a lot longer (and more were made) so you would expect them to have more incidents. But does that mean they are safer? Less safe? Obviously it's all in fun, but the numbers would make for some interesting discussions on the forums.

Obviously the 12s are way safer than the others ;-)
 
John B - funny that you should create this I started doing it in excel based on a post a few days ago. What might be fun is to see the "A" vs tail wheel comparisons and also the % of incidents per model as a percentage of models flying from Vans website. The 6's have been around a lot longer (and more were made) so you would expect them to have more incidents. But does that mean they are safer? Less safe? Obviously it's all in fun, but the numbers would make for some interesting discussions on the forums.

Obviously the 12s are way safer than the others ;-)

Yeah, I could probably add in the fleet sizes and do some more work there to make the numbers more meaningful. However, knowing how messy the "make" and "model" fields of the dataset are, doing a tail v nose wheel comparison will probably be painful. Not only do you have the builders registering interesting model names, you also have the NTSB typing them in an inconsistent format, as I discovered when running a similar report on Mooneys. As with anything human entered, you get all kinds of funny stuff.
 
That analysis tool is way cool, but it's important to realize the limitations of the underlying data.

For the EAA Founder’s Innovation Prize, I analyzed 551 RV-series reports in depth, meaning I had my own data-driven taxonomy implemented on a multi-page spreadsheet and went through the entire data set multiple times, a tedious and exhausting exercise. With regard to flight phase, I found that the flight phase information was often missing, irrelevant to the accident chain, or inconsistent with the narrative. Not the NTSB’s fault, they aren’t necessarily the ones whose staff enters and edits the data. But you get the idea…

What my research (which included way more than just the NTSB reports) showed was pretty interesting, and it did not support the hypothesis that giving everybody an angle of attack display would solve loss of control problems. Rather, the two main conclusions were:
1) Safety is not part of the general aviation culture in the US. In the experimental amateur built aircraft community, the culture is, “It’s experimental, you can do what you want,” ignoring that the laws of aerodynamics don’t care what you want. If safety is to be part of the culture, it must omnipresent – in every magazine article, not just in a safety column (which emphasizes unintentionally that safety is separable from all the good stuff); in every award (e.g., EAA homebuilt judging doesn’t require that a full, documented flight test program be performed, only that the plane is flying); at every EAA chapter meeting (videos from the mother ship would be easy to do); in all other communications and meetings; and most importantly, the underlying theme of those safety messages should be, hey, we value you and want to keep you around, not the current, limited effectiveness message of look what yet another lamentable underachiever (*******) did.
2) Basic stick and rudder skills are frequently inadequate, often because conventional flight training focuses on proficiency in the middle of the envelope – accidents don’t happen in the middle of the envelope – with occasional, timid forays towards stalls. What’s absent is any training or documented exercises to improve flying skills in that great gap between middle of the envelope and stalls, let alone spins. I’ve developed a whole slew of exercises that are fun, challenging (tried on both ATP and private pilots), and those exercises also give pilots something fun to do besides fly around in circles. These are at most 2 G maneuvers that should be doable in just about any GA airplane, and I’ve tried these so far in RV-9A, RV-8, and Cessna 150. The point is to help people recognize, avoid, and survive potential LOC situations.

The Chairman of the NTSB requested the best copy of my work, and he held a meeting last week to review that work and other material. In an hour long, follow up phone call with a staff member, my work was called, “thought provoking.”

Now it’s time for the next step, to get the greater aviation community involved so that these ideas mature into our ideas instead of just Ed’s ideas. If anybody has any contacts with the aeronautical universities (ERAU, UND, LeTourneau, etc.) or any of the big flight schools, I’d love to work with them, and I’ve got all the appropriate credentials. I think that at this stage, having an organization work with and develop the concepts would be more efficient that working with individuals – which would probably be the next step.
 
Last edited:
Ed, why don't you post your exercises for us to evaluate?

Thought about that, but...
* Way too many pages to post, and the present descriptions of the exercises are adequate but terse;
* Some of them really should be done with a CFI who has seen those exercises in advance and knows how to do them -- not because they're dangerous, but because they're so different. Some of the exercises deliberately generate high levels of sensory input that could task-saturate the uninitiated;
* At this time, the objective is to mature the concepts and make them mainstream. There's a perceived risk to that objective if the exercises are just thrown over the fence and one or two people loudly trash them, whether out of ignorance or self-aggrandizement. That's always a risk with new ideas;
* I don't have any good way to vet the volunteers, especially in terms of their ability to generate meaningful, thoughtful, insightful comments;
* Everybody's favorite, liability.

Thanks for the thought, though.

Ed
 
Last edited:
Great Data-Change of Habits

Does this data take into account flight hours or fleet sizes? If not, what value does this data provide?
Been reading just the NTSB reports for the RV-6 series.

A few are unknown issues, but two issues have struck home for me:

Open canopy in flight and Carb Icing.

Just how does an open canopy in flight happen? Well it looks like on take-off.

This time of year here in Alaska the canopy frosts over instantly if you have been doing any physical activity to pre-flight and pre-heat. Removing covers, checking and removing any airframe ice, warms you up fast because of layers to stay warm. Once you climb inside and shut the canopy for engine start, instant frost that starts on your side of the canopy.

My answer has been to open the outside vents, letting the cold air in to circulate, then opening the canopy so it vents to the outside while I taxi for takeoff. This keeps the canopy clear, once ready for run-up close canopy latches, do run-up and go. Could it be possible to forget to latch the canopy?-yes almost did it, after taking the active runway, realized it was still open.

Change of habit, bought a small 12v electric "defroster" from Walmart. Not sure if it works, but if it can circulate the air enough without using the heat option, it can be a life-saver. Won't be taxing with the canopy open anymore!

It looks like the canopy wasn't latched on a few of the accidents possibly due to high heat and ventilation while taxing.

Headsets are great, but they also lessen the noise from an open canopy. Can't quite here the traffic or tower, turn the volume up...

Carb ice seems to be another issue, from now on I'll do a Carb Ice Potential Chart check before every flight. Interesting to see it can happen with certain conditions above 70 deg F. Rarely gets that hot here in Alaska, so need to be more vigilant when pre-flighting.

Great info if you want to learn from other's mistakes. Well worth the time to read!

Best regards,
Mike Bauer
 
Been reading just the NTSB reports for the RV-6 series.

A few are unknown issues, but two issues have struck home for me:

Open canopy in flight and Carb Icing.

<snip>

Best regards,
Mike Bauer

My rule for the canopy (slider) is that it is either wide open or it is closed and latched. It is never shut but unlatched.
 
Carb ice seems to be another issue, from now on I'll do a Carb Ice Potential Chart check before every flight. Interesting to see it can happen with certain conditions above 70 deg F.

I found a significant number of carb ice events in the NTSB data. I get some RPM drop in my RV-9A, but the carb heat worked well the one time I needed it in flight. In the RV-8, I see no RPM drop during runup. And in a friend's RV-9A, applying the carb hear *drops* the carb air temperature one degree, according to his instrumentation.

Ed
 
My rule for the canopy (slider) is that it is either wide open or it is closed and latched. It is never shut but unlatched.

Basically the same for me in the tip-up, it's either open, resting on 'hook' or closed latched. Let's just say I've learned this lesson the hard way.

I can also tell you that flying a 7 with the canopy unlatched won't kill you, unless you forget to fly the airplane.
 
Back
Top