Mµ Thermal Imager - real or fake?

[BulletMagnet83]

Yeah... after someone helpfully replied to my question and told me the cheapest sample unit they'd found was $2000, it does leave me somewhat deflated (because I can't have one  ) and wondering if Mu even HAD an imaging sensor until they'd already received the backing money.  I mean, you wouldn't buy just ONE to prototype with, would you? You'd need at least one spare. The fuck-up fairies demand a sacrifice after all, and if one is all you have, it'll blow up or burn out.

[Corporate666]

I think the Mu guys did some very (very) basic costing from a perspective that was absolutely naive about manufacturing and production.  A lot of people put $$ in, and I'll bet their money supply has been eroding ever since.  They bought hardware and software, prototype parts, plus the cost of just running a business, even "on the cheap".

They certainly aren't going with all the cheapest parts (i.e. the Sam3x and Redpine module), so I think what will happen is they'll get to the point where they think they can make something... only to find out they have only 1/5th of the funds needed to do a production run.

That will be the oh-shit moment.

[mikeselectricstuff]

I think the Mu guys did some very (very) basic costing from a perspective that was absolutely naive about manufacturing and production. 

..and clueless about the physics and costs of thermal imaging - I bet they thought it was little different to normal image sensors and optics.

A lot of people put $$ in, and I'll bet their money supply has been eroding ever since.  They bought hardware and software, prototype parts, plus the cost of just running a business, even "on the cheap".

They certainly aren't going with all the cheapest parts (i.e. the Sam3x and Redpine module), so I think what will happen is they'll get to the point where they think they can make something... only to find out they have only 1/5th of the funds needed to do a production run.

That will be the oh-shit moment.

I can't believe that at this stage they don't  know how much the sensor and optics will cost. That would have been the Oh Shit moment...

[MFX]

I can't believe that at this stage they don't  know how much the sensor and optics will cost. That would have been the Oh Shit moment...

Yes, surely if they actually have a sensor and optics they must know the cost. Unless they have some sort of shady deal to get substandard stuff out of the back door of some dodgy Chinese factory then how are they getting a sensor/optics at anywhere near the cost needed to stay in budget.

[AlfBaz]

Unless they don't know the difference between near and far infrared and thought the could just remove the ir filter off a standard sensor

[tom66]

I wonder if it's possible, perhaps even considerably cheaper, to buy thermal imaging sensors with many dead pixels? I can buy cheap China PAL/NTSC cameras for about £4 from eBay, shipped from Hong Kong. They typically have around 50~100 dead pixels on a 320x240 array. But they're cheap, so I can tolerate that. Not sure if I'd like it so much with a relatively expensive thermal imaging camera, but it could work as a compromise to achieve low cost.

[mikeselectricstuff]

I wonder if it's possible, perhaps even considerably cheaper, to buy thermal imaging sensors with many dead pixels? I can buy cheap China PAL/NTSC cameras for about £4 from eBay, shipped from Hong Kong. They typically have around 50~100 dead pixels on a 320x240 array. But they're cheap, so I can tolerate that. Not sure if I'd like it so much with a relatively expensive thermal imaging camera, but it could work as a compromise to achieve low cost.

I would guess a significant cost is the (vacuum) packaging and germanium window. The pixels are quite big, and the calibration shutter allows for fixed-pattern correction of some pixel-to-pixel variations, so it may be that yields are already quite high, so I'd be surprised if this would be a major saving.

[CanadianAvenger]

It's been few weeks since they posted any update, and the natives are getting restless again. They were so close to posting some real, tangible proof, but then they stopped short... I don't get it. These guys are way out of their league, both in terms of running a crowd funded campaign, and actually developing a product.

[Rasz]

These guys are way out of their league, both in terms of running a crowd funded campaign, and actually developing a product.

I dont know, as far as scams go they are doing great.

[Recyclojunk64]

Another update. It would seem that they're still working on all the bells and whistles before actually doing the real work on the actual sensor and processing.

Hello All,

So right now we're tweaking and modifying the board's firmware to make the most stable and consistent platform possible for the release product. What we arire aiming for is the fastest possible startup and warm-up for the wifi network and the imaging sensor respectively, as is possible. This process involves a considerable amount of testing and trials.

We have been in design and manufacturing discussions with several case manufacturers over the last few months, however, we are still willing to consider new providers. To that end, if any of you out there are camera afficianados and have a camera case manufacturer that you really love, please let us know in the comments, and we will look into what they have to offer.

As for the camera body itself, while we continue to work with the hardware, we are using the opportunity to improve the feel of the device. If there are any features that you would like to see, please let us know.

We are quickly approaching the point where we feel we will be able to provide a new release date and look forward to doing so.

Thank you for your continued support,

-Mu

[Rasz]

Hello All,

So right now we're tweaking and modifying the board's firmware to make the most stable and consistent platform possible for the release product.

translation: our firmware crashes constantly

What we arire aiming for is the fastest possible startup and warm-up

translation: our highly optimized java code rewritten into arduino sketch boots in 2 minutes

for the wifi network and the imaging sensor respectively

translation: wifi nor imaging sensor dont work at all

To that end, if any of you out there are camera afficianados and have a camera case manufacturer that you really love, please let us know in the comments, and we will look into what they have to offer.

translation: pleaser do it for us for free

We are quickly approaching the point where we feel we will be able to provide a new release date and look forward to doing so.

translation: we are running out of excuses

[EEVblog]

We are quickly approaching the point where we feel we will be able to provide a new release date and look forward to doing so.

But still no thermal image 

[senso]

That afficcionados, they must be watching daves movies 

[Bored@Work]

We have been in design and manufacturing discussions with several case manufacturers over the last few months, however, we are still willing to consider new providers. To that end, if any of you out there are camera afficianados and have a camera case manufacturer that you really love, please let us know in the comments, and we will look into what they have to offer.

I lost track of all the claims. Didn't they once claim they had a manufacturer, and they were expecting to get prototype enclosures any day?

[baljemmett]

We have been in design and manufacturing discussions with several case manufacturers over the last few months, however, we are still willing to consider new providers. To that end, if any of you out there are camera afficianados and have a camera case manufacturer that you really love, please let us know in the comments, and we will look into what they have to offer.

I lost track of all the claims. Didn't they once claim they had a manufacturer, and they were expecting to get prototype enclosures any day?

For enclosures, yes -- I seem to recall they've claimed several times to have a prototype enclosure ready to go, just in time to redesign the entire thing to use another handy COTS sensor that nobody else has heard of, etc.  However, I read that paragraph as referring to a soft or rigid carry case or similar -- hence the plea for 'camera afficianados [sic]' to recommend their favourite manufacturers.  Presumably not many camera geeks would have useful suggestions to make regarding the actual device enclosure.

[Corporate666]

Another useless update - they are trying to decide about banana flambe or creme brulee for dessert when they don't even have the beef roast cooking yet for the main course, nor do they have anything to cook with.

It seems pretty clear at this point that they showed some off the shelf stuff figuring that will get them the $$$, and the $$$ will let them design the product.

The interesting thing - the microbolometer - remains a mystery.  And will continue to remain a mystery to the end (the end being 99.9% probability of implosion, and 0.1% probability of delivering something).

[EEVblog]

I lost track of all the claims. Didn't they once claim they had a manufacturer, and they were expecting to get prototype enclosures any day?

I think the biggest claim is that they are using an off-the-shelf sensor that "has been on the market for some time".
Given that that was many months ago + "some time" + NDA type pre-release time for those in the the know, that an awful long time for this magical sensor to have been available without anyone having ever made a cheap 160x120 camera from it. Even if the sensor alone was the asking price of this Mu camera, there would be half a dozen of them on Hack-a-Day by this point. Yet we see zip, and no one in the industry seems to be aware of this sensor...
And of course, the sensor is everything.

In the words of an infamous aussie "please explain..."

[CanadianAvenger]

Latest update posted last night:

Hello All,

We've got the portion of the firmware that deals with the wifi all set, the camera starts up and is ready in just over 2 seconds. We're very happy with this startup time, and better yet, we have a very solid and stable throughput on the new board.

A few more tweaks to the portion of the MCU firmware that deals with the ADC and the sensor itself and the firmware should be good to go.

These are major advances and moves us closer to our final goals.

We'll keep you updated.

-Cheers

"The firmware should be good to go"...  So that means they have images coming from the sensor right?  So why the hell have they not posted any?!? [Don't answer, I already know ]

[Corporate666]

"The firmware should be good to go"...  So that means they have images coming from the sensor right?  So why the hell have they not posted any?!? [Don't answer, I already know ]

The chef emerges from the kitchen to address the staving diners who have been waiting 3 hours for their meal, and speaketh...

"Don't worry folks, we have managed to get the texture of the glaze on the dessert fruit cup we're serving for dessert just perfect, you will really enjoy it!"

but the diners just want their damn fish and chips main course!

[thermalguy]

Hello all.

It took about 2 weeks longer than i wanted but I have modified one of our production 320x240 bolometer units to deliver 80x60 via USB to a Java application running on windows. This was accomplished by working a couple of hours in the evenings. My plan to use bluetooth fell apart when I blew up one of the demo boards. :| That is what happens when you don't pay attention to what power supply you are hooked up to. So I went with a USB board from our eval kit. Didn't have to spend any money though.

I have attached an image of the modified core as well as an image from the data stream. I left the thermal image in its native resolution. Pictures are great, but video is better so I am going to create a video of the unit streaming this week and post it to our youtube channel. Given our present hardware design, we would have to remove MANY of the parts on our current design and shrink the package down. That is the only way I see to get the sales price down in the $500 range and that would be with 2000 people signing up to take units.

Is there any interest at that price point?

To extract the data I see backend daughter boards being used.  You select one when you buy the unit and you can come back and order other ones if you want. Among the list are USB, Wifi, and Bluetooth. However, if we can generate more interest I am sure this would be a good kickstarter project for us.

Couple of notes on the IGG Mu Optics project.

1. Before I did any of the work above I contacted them an offered my help to get their project completed. No one responded AT ALL.
2. I did not see any external memory chip on the picture they posted. Maybe its underneath the board. If not, there is no way they can do any frame processing with the SAM3X; it doesn't have the resources. Line processing maybe, but you are going to push the heck out of the DMA controller.
3. If they run non-temp stablized (best for battery operated scenarios), they will need image NUC data for the whole array across whatever ambient operating temperature profile they select. That eats up alot of flash. I did not see an extern flash memory chip either.
4. In order to measure temperature with accuracy, you need to know what the lens temperature is. Low cost methods couple the lens temperature with the detector temp. Since the lens mount looked 3D printed, the approached wasn't utilized.

[mikeselectricstuff]

... That is the only way I see to get the sales price down in the $500 range and that would be with 2000 people signing up to take units.

Is there any interest at that price point?

That's what Kickstarter is for.... to find out if there's a market
Not sure $500 quite brings it into the 'cool toy' range, but at half the cost of a Flir i3 could be interesting, especially if it offers features the Flir doesn't - e.g. image storage to your phone.
I wonder if there would be useful mileage in open sourcing the firmware so people could add their own functionality. if nothing else it would be good for free publicity ( and just ignore the OSH fanatics who will inevitably criticise for not open-sourcing the hardware, the bolometer design and everything else..)

To extract the data I see backend daughter boards being used.  You select one when you buy the unit and you can come back and order other ones if you want. Among the list are USB, Wifi, and Bluetooth.

USB pretty much comes free with most microcontrollers, and provides your charging connector and a means to update firmware, so I can't see a reason for this not to be the minimum standard interface.
SD card interface is also negligible hardware cost  so probably not worth omitting.

Couple of notes on the IGG Mu Optics project.

1. Before I did any of the work above I contacted them an offered my help to get their project completed. No one responded AT ALL.

No surprise - they probably thought it was a competitor trying to sabotage them, or still believe they need no help.

3. If they run non-temp stablized (best for battery operated scenarios), they will need image NUC data for the whole array across whatever ambient operating temperature profile they select. That eats up a lot of flash. I did not see an extern flash memory chip either.

Serial flash costs peanuts these days. Probably plenty of scope for compression & interpolation to save space.

[EEVblog]

I have attached an image of the modified core as well as an image from the data stream. I left the thermal image in its native resolution. Pictures are great, but video is better so I am going to create a video of the unit streaming this week and post it to our youtube channel. Given our present hardware design, we would have to remove MANY of the parts on our current design and shrink the package down. That is the only way I see to get the sales price down in the $500 range and that would be with 2000 people signing up to take units.

Is there any interest at that price point?

The best way to judge it would be to run a crowd funded campaign and see. I think it's quite possible you could get 2000 people at 500 a pop.
Just make sure you really show of that proto and what you can do and that you are an existing professional manufacturer.
Although remember that the Flir i3 is under $1K already. But geeks do like toys for their iPhone.

[ElectroIrradiator]

I took the liberty of scaling the thermal image 6 times, with zero interpolation. At this size it is probably a bit easier to see what is going on, given modern monitor resolutions. Hope thermalguy doesn't mind too much.

Looks like it would be useful for detecting hotspots on PCBs and suchlike.

[Rasz]

more like this, there is no reason to not at least do some post

[ElectroIrradiator]

more like this, there is no reason to not at least do some post

As long as you remember that any added details are artificial constructs...