[*sean*]

I know this is WAY out of scope, but a feature request: a Stack Overflow-like section for Insight questions. Yeah, totally different thing. But seeing the word "posts", I thought of this, something I wish for every time I am searching a huge thread for some arcane tidbit of information.

Oh, and this is AWESOME by the way! And a tip of the hat to @julianachos for all the manual verification!!!!!

 

[AfterEffect]

Finally ready for full release! Check out the registry here:

https://www.tinkergear.com/insight/dashboard.html

To update your car in the registry, search your VIN on the dashboard or go here:

https://tinkergear.com/insight/register_entry.html

Please let me know if anything breaks or if you have any suggestions.

EDIT: Canadian market VINs are now supported, courtesy of @julianachos

 

[Gasoline Fumes]

Should the registry update automatically? Mine doesn't update with the submitted info.
JHMZE1377YT005342

Edit: Never mind! I see that it will be reviewed first! 🙃

 

[AfterEffect]

Should the registry update automatically? Mine doesn't update with the submitted info.
JHMZE1377YT005342

Just approved the pending entries. All of them run through a review process, since there is a significant risk of bots or trolls poisoning the data or uploading questionable material with how I have it configured. I didn't want to require users to create an account or share their personal information just to be able to update their VIN.

 

[Gasoline Fumes]

Just approved the pending entries. All of them run through a review process, since there is a significant risk of bots or trolls poisoning the data or uploading questionable material with how I have it configured. I didn't want to require users to create an account or share their personal information just to be able to update their VIN.

Sorry for entering mine 3 times!

I'll update my other Insights later. I'm away from home.

 

[rodthies]

I have compiled a comprehensive list of Gen1 Insight production VINs and I am now confident enough in the data to release it.

This data was collected by first generating a list of all possible Insight VIN numbers using the US sales estimates, and adding 1000 units as a buffer in case there were large gaps. This produced ~38,000 possible VINs.

This list was then checked against Hondas official database to confirm whether it was a real VIN and collect other information about it.

Once I had a list of real production VIN numbers, the list was checked against Carfax/NMVTIS data to verify it. If there were Carfax entries, that was confirmation that the VIN was real and was sold in the USA.

I am now in the process of running this full VIN list through Google via their Cloud API. Hits are being recorded and categorized by website. This will produce a partial list of cars that have been wrecked and likely crushed, and fill in some other additional fields.

I am not running a full Carfax report on each VIN due to the cost, but I am building a registry portal that will be hosted on my website. It will display fancy graphs of the data, and allow users to enter VINs of their own cars or others they know the status of.

I will not be releasing the full database of VINs due to potential errors or privacy issues, but if there are any suggestions for the portal, please don't hesitate to ask.

With all that out of the way, here is what I consider to be the complete and authoritative list of Insight production numbers in the US:

2000 - 5,554 (All MT, 2538 A/C, 3016 Non-A/C, 3736 Silver, 1355 Red, 463 Citrus)
2001 - 3,408 (1908 MT, 1500 CVT, 1581 Silver, 639 Red, 1188 Blue)
2002 - 2,001 (633 MT, 1368 CVT, 1244 Silver, 279 Red, 478 Blue)
2003 - 1,010 (334 MT, 676 CVT, 659 Silver, 121 Red, 230 Blue)
2004 - 535 (157 MT, 378 CVT, 340 Silver, 71 Red, 124 Navy)
2005 - 578 (257 MT, 321 CVT, 309 Silver, 146 Red, 123 Navy)
2006 - 802 (447 MT, 355 CVT, 359 Silver, 202 Red, 241 Navy)



*Disclosure - While I believe the data to be 99.9% accurate, it is subject to change. Link to registry to be added soon.

thanks for compiling all this info, Mike. very helpful! i'm the recently joined member who is the original owner of the '06 Silverstone Metallic CVT w/56K miles on it w/serial #678. when we spoke on the phone you mentioned Honda may have had some "gaps" in their vin #s if i understood you correctly. do you have any insight( sorry, LOL) as to why honda had these gaps? Britney mentioned you might be getting a canadian aquamarine car. if so, what do you know of its history & maintenance & is it a nice, running car?

 

[AfterEffect]

you mentioned Honda may have had some "gaps" in their vin #s if i understood you correctly. do you have any insight( sorry, LOL) as to why honda had these gaps?

Most of the gaps were early vehicles in 2000, likely test/crash cars. There are some other small gaps, which seem to be cars that were produced, but were not sold for one reason or another. Julian may know more about this than I do.

 

[Aerocool]

My takeaway from the website was that 2001 was the year of the thunderdome for the Honda Insight.

Baseless speculation: unidentified eccentric oil tycoon billionaire and prolific hybrid car hater buys majority of available US stock in 2001 to host the world's first private hybrid car destruction derby.

 

[Natalya]

Ask a Honda technician to run JHMZE14751T000001 in one of their systems (I think this one in particular was under an internal parts catalog) and it will show up. There was no history of it being sold, though. I had a friend who was willing to help me on this project to find certain cars to determine whether they existed or not.

Canadian 2001 with VIN 800001 is SawBite, which Honda gave to the Canadian government for testing, and was never sold. So, maybe something similar was done with American 2000 VIN 000001?

 

[mudder]

Super awesome, great to finally see this data all in one place.

mods, I nominate this post for a sticky.

Please let me know if anything breaks or if you have any suggestions.

Add an option for "I don't know mileage". This will help us document junkyard cars without random "123456" or "999999" mileage entries.

I feel like the registry update page should auto-populate with any given VIN's existing entry data.

I'd love to see all the public VIN info available publicly. Or should we just scrape your website? .

Would be neat to see the "Registry by Color" and related circle graphs update dynamically based on model year selection.

Can we add a "license plate" entry on the registry update page...
...and then eventually a "search by license plate" feature? Would go well with Pete's "I saw an Insight at..." post.

I didn't want to require users to create an account or share their personal information just to be able to update their VIN.

For LiBCM, I've managed to prevent bot submissions by asking the question "what is this site's owner's insightcentral.net handle?" And then hyperlink it to your profile.

...

Thoughts on the data:
-I (incorrectly) thought that silver and red production numbers were approximately equal, whereas this dataset shows a 3:1 ratio between silver:red.
-I'm amazed that nearly 1 in 4 cars didn't have AC.
-Do we really only expect that QTY6010 G1 Insights are still roadworthy? How is this estimate calculated? If this number is close, then there are ~QTY50000 non-Insights on the road for every lonely G1 Honda Insight. Sad 52 Hz whale tone!
-WOW, production is only QTY3 on MY2003 Canadian Insights!
-Am I reading the data correctly that ALL Canadian Insights were MT... and had air conditioning? That's what the bar crash suggests. Seems wrong.
-My daily driver JHMZE13581T001563 is marked "without A/C", but I believe it has the factory installed A/C. How do I tell if my MY2001 AC is factory or dealer installed?

...

How 'useful' would blanket CarFax data actually be for every VIN?
Would this cost a few thousand dollars?

 

[Natalya]

I've never seen a Canadian CVT Insight. Manuals are more popular in Canada than in the States so it would make sense Honda didn't offer the CVT there.

 

[AfterEffect]

Add an option for "I don't know mileage". This will help us document junkyard cars without random "123456" or "999999" mileage entries.

Good ideas, thank you.

The intent was to view the VIN information on the search page, and then navigate to the entry page if it needed updates. Any fields left blank retain the last value. Do you think that the fields should auto populate on the entry page as well?

More data will be available "soon". A lot of the graphs that would be nice to have will be incorrect until more data is gathered. You're right that the colors should update when selecting a year.


I was also shocked at the number of non-AC cars. I have seen a number of '00 Insights without AC, but never any other years. I have a number of entries claiming that their cars have factory AC despite the VIN decoding as non-AC, which makes me wonder how many were dealer converted.

The ~6010 Insights is some napkin math and magic numbers, I will admit. It uses production by year * average attrition rate by year * number of years to present with a small amount of randomization. I manually reduced the average attrition rate slightly under the assumption that these are enthusiast cars and may be slightly less likely to be crushed than your average Civic or Corolla. I compared the resulting numbers to my scraped data of average number of Insights that have run through the insurance auction per year, and it was pretty close. I hope to have more accurate numbers once I integrate that data and get more user submissions.

All the Canadian data was collected by Julian, so I have no data to verify it against. His USDM data was pretty much completely in agreement with mine though, so I have no reason to doubt those Canadian numbers.

 

[julianachos]

when we spoke on the phone you mentioned Honda may have had some "gaps" in their vin #s if i understood you correctly. do you have any insight( sorry, LOL) as to why honda had these gaps?

Most of the gaps were early vehicles in 2000, likely test/crash cars. There are some other small gaps, which seem to be cars that were produced, but were not sold for one reason or another. Julian may know more about this than I do.

I'm not speaking for Honda on this one, but I will say this about what I have learned outside of Honda. Don't ask me the details on this one - I know it sounds very "trust me bro." I will, however, explain some things below related to this project, since it was done with mostly public information.

My assumption for the 2000MY gaps (especially the huge ones from YT000001 through YT000043) is that they were either test cars meant to stay close to American Honda, were then immediately scrapped and/or were once pre-production press cars. For all we know, there might be a VIN for that car, but there is no way to find out unless you had internal access in 1999 when everything was in the moment. There really is no rhyme or reason behind this bit, as it probably wasn't intended to be questioned.*

The smaller gaps that AfterEffect mentions might be cars that were pulled off the production line for either testing or QC purposes and were never sold, or for special occasions. I once saw a vehicle at a Toronto auto show, and a year later, I personally helped dismantle that very same car in the States because it was a vehicle marked as a pre-production model. With that said, it was requested by the manufacturer to be scrapped and never sold to the public. I remember the VIN not being anything special, such as a low serial number, which I've seen multiple pre-production cars in similar positions before.

The VINs of the cars that did appear through reverse engineering and some method of verification or digital bread crumb have been scrapped during the production years were likely pre-production press cars if I had to guess. These are the cars that did appear in the YT000001 through YT000043 gap.**

*Here is some evidence that backs up my statement, since this was all done in a public-ish setting.

Below is a photo of a 2021 Acura RDX PMC Edition that was parked outside of Acura PMC and its VIN tag.

If you were to run that VIN on Carfax and Acura's MyGarage portal, nothing would show up. When you find a standard '21 RDX PMC, like this one for sale, you'll notice the VIN is a little different. Searching the VIN on the two websites mentioned will generate a result.

The 9th position (green) is a check digit, which the orange car has the correct calculation. What sticks out for me is the 11th, 13th, and 14th position.

For the 11th position, L designates the East Liberty, Ohio plant, which is around the corner from the Marysville plant. Y is seen on second-generation NSXs like this one. From my quick passive searching, I've only seen RDX PMCs with L in the 11th position.

The 13th and 14th position pattern from the orange car was not found in other RDX PMCs I've seen for sale.

TLDR: VINs that return zero results may actually exist, but are meant for some sort of internal use only.

** Here is another example, but with results that would appear. This car is also under Acura's possession.

Interestingly enough, this particular NSX resembles a Type-S, which was released in 2022, even though the VIN shows it is a 2017 model. For all we know, maybe the dashboard was swapped in to make the car complete - just speculating here.

With the above statements said and again, things like this just kind of exist with no real logic behind VIN patterns, etc. with pre-production cars if there is a VIN result. I have photos of another NSX in a similar fashion, but with an odd VIN like the RDX. That one sports a 2019MY VIN, which returns no results on Carfax and Acura's portal.

TLDR: The whole thing about the pre-production cars is a weird gray area. The random VINs I found such as 1T000001 that AfterEffect couldn't find, probably aren't counted in the sales figure that Honda reported. Do note I managed to uncover more VINs than the sales figure.

I know these cars fall outside the gap mentioned above, but hear me out on my rambling for the sake of science. According to my notes, Insight YT000050 generated four Carfax reports. On a similar website, there was a note indicating it was a corporate fleet vehicle, which later received a salvage title in 2003 - it is likely scrapped. YT000059 was featured at the 2000 North American International Auto Show, which has a Monroney label and one Carfax history report that mentions AHM.

 

[julianachos]

Canadian 2001 with VIN 800001 is SawBite, which Honda gave to the Canadian government for testing, and was never sold. So, maybe something similar was done with American 2000 VIN 000001?

Most of the gaps were early vehicles in 2000, likely test/crash cars.

YT000333 has three Carfax reports, which initially was a car I had no issue finding. This one is interesting, though. It was sold to a government contracting company called KARCO Engineering, which performed the crash testing for NHTSA. With that said, I don't think this explains the gaps such as 1T000001 or even YT000001.

1T800001 (Sawbite) has nine Carfax reports. AfterEffect can attest to this since he is the only other person with that data, but I found 409 VINs for Canada. The reported sales figure was 399. Those ten cars generated a VIN despite not appearing on Carfax, however they were found through an internal source.

Sawbite was not one of those ten cars.

I'd love to see all the public VIN info available publicly. Or should we just scrape your website? .

@AfterEffect: he might be an exception in our one conversation earlier regarding security. Good luck.

Maybe when I visit and you ask nicely, I will show you the findings. I showed a few people at the Frederick meet and InsightFest last year. I am waiting for AfterEffect to verify my manually found data using a script, since he did find one incorrect VIN on my sheet, to my knowledge at the time of posting this.

Thoughts on the data:
-Do we really only expect that QTY6010 G1 Insights are still roadworthy? How is this estimate calculated? If this number is close, then there are ~QTY50000 non-Insights on the road for every lonely G1 Honda Insight. Sad 52 Hz whale tone!

I mentioned FOIA requests in post 14 that I was originally going to do; however, I decided to hand the torch to AfterEffect to deal with all of that. I hate to rain on the parade here going, "wELL aCkcHyuALLy," but that data supplied by the DMVs/MVAs/RMVs/(insert whatever acronym here) will only show what is registered, not what is truly out in the wild - similar to the UK-based website "How Many Left." I'm sure we can think of a few people here who have Insights that could theoretically be on the road, but aren't.

-Am I reading the data correctly that ALL Canadian Insights were MT... and had air conditioning?

Correct.

-My daily driver JHMZE13581T001563 is marked "without A/C", but I believe it has the factory installed A/C. How do I tell if my MY2001 AC is factory or dealer installed?

Under the hood somewhere should bear a sticker that looks like this, added by a dealer technician. I pulled it off a car at the junkyard that had a 5 in the 8th position of the VIN.

The VIN alone is a clear-cut answer to the installation method, full stop.

How 'useful' would blanket CarFax data actually be for every VIN?
Would this cost a few thousand dollars?

In my honest opinion, blanket Carfax data would be great to determine what Insights have recent (~3 years) reports to determine what is on the road and what isn't. I was using Orto in the early days of this project and found a number of 2000MY cars that had recent reports at the time. Orto was an app that generated reports just like Carfax. I ran a few VINs on both Orto and Carfax at the time, and both produced the same data. It cost me a one-time fee of something along the lines of $9.99 for unlimited entries, but it stopped working after a few months of me paying for it.

I got as far as YT000352, which means I found 126 cars that, in theory, were driving at the time. This was using the ~3-year criteria.

To keep the data consistent, reports will likely need to be run on the cars with said recent data to see if/when they fall off. The other fault I see in this is what if a car that was once brushed off, thinking it was done, decides to come back from the dead? I can't genuinely answer whether it'd be worth the hassle.

Here are the prices for the general public:

I believe dealers have the option to have an unlimited plan, but I'm sure the cost isn't cheap. I think the plan is on a month-to-month basis, but don't quote me on it.

 

[*sean*]

@AfterEffect or @julianachos

Does the serial number and CVT vs MT (or for that matter, whether or not it has AC, or the color) match any pattern like "today/this month we are making CVTs" or "today all cars are red"?

Does this differ at the beginning of the model year?

Is there a VIN 0001 for 2004? I have 0002, and there is evidence suggesting that it was used for tuning or testing before it was sold.

What is the VIN of the first MT of 2004?

And finally, what is the VIN of the first US CVT in 2001? Were any MTs made after that CVT in 2004? (This last one is to answer a question I've had for a long time: why does the Electrical Troubleshooting Manual discern list "late 2001" for changes in the air fuel ratio sensor? All CVTs (AFAIK) use a Denso four-wire sensor, while MTs from 2002 on use the Denso (a rectangular black module inside the ECM that handles the AFR sensor is also marked Denso; in the 00-01 MTs it is a green potted module.) Wondering if the changeover to Denso for MTs might have happened in late 2001 for a few cars or if "late 2001" really means "2001 CVT".)

Finally finally, many people with project or parts cars (like me) won't be able to update the registry since mileage is a required field and the car requires a working 12V battery or an instrument cluster to read the mileage. (A "rough estimate" checkbox for those???)

 

[julianachos]

Does the serial number and CVT vs MT (or for that matter, whether or not it has AC, or the color) match any pattern like "today/this month we are making CVTs" or "today all cars are red"?

Does this differ at the beginning of the model year?

I noticed patterns of one color being made, then having various configurations of manual/base, manual/AC, and/or CVT, if that makes any sense. The serial number had no deciding factor from what I have seen. There were some stretches of manual/base, manual/AC, and/or CVT in various colors, but the colors seem to come first.

To directly answer your question, my guess is they would go, "We are making all red cars today, but mix them up."

At the beginning of the model years, it tends to follow the above. Sometimes the string of one color only has one 7th and 8th position VIN combo.

Is there a VIN 0001 for 2004? I have 0002, and there is evidence suggesting that it was used for tuning or testing before it was sold.

There is no VIN for 4T000001.

What is the VIN of the first MT of 2004?

4T000007.

And finally, what is the VIN of the first US CVT in 2001? Were any MTs made after that CVT in 2004? (This last one is to answer a question I've had for a long time: why does the Electrical Troubleshooting Manual discern list "late 2001" for changes in the air fuel ratio sensor? All CVTs (AFAIK) use a Denso four-wire sensor, while MTs from 2002 on use the Denso (a rectangular black module inside the ECM that handles the AFR sensor is also marked Denso; in the 00-01 MTs it is a green potted module.) Wondering if the changeover to Denso for MTs might have happened in late 2001 for a few cars or if "late 2001" really means "2001 CVT".)

See posts 3 and 8.

 

[*sean*]

See posts 3 and 8.

I asked this less than a week ago? I am losing my mind.

Thanks for the info. I guess I have #1 from the 2004 model year. And it is a rare color, Navy Blue. Sadly, it lived in Rustvillle USA previously. I pulled the transmission to replace the urea grease and check the splines. I better put it back together!

 

[UK_MJN]

Great stuff! Regarding your observation of gaps in the 2000MY VINs, you may find some of these when you get to EU or JP cars? We have JHMZE1350YT000032 (has A/C fitted) but it fails your check digit test. It was first owned by Honda UK and went to JOHN WILDING HONDA in Kendal, Cumbria.

 

[julianachos]

Great stuff! Regarding your observation of gaps in the 2000MY VINs, you may find some of these when you get to EU or JP cars? We have JHMZE1350YT000032 (has A/C fitted) but it fails your check digit test. It was first owned by Honda UK and went to JOHN WILDING HONDA in Kendal, Cumbria.

I don’t think either of us will be able to get an extensive list with those two markets like we did with the North American cars, however I mentioned to AfterEffect that it was a good idea to collect the data, even though his registry currently does not support it.

Only the North American cars have the check digit in the VIN. This is the way with many cars, if not all.

 

[UK_MJN]

Thanks. Odd that the check digit isn't global! Be very happy to help collect the data. Note that we don't have a build date on the door pillar either.