Mµ Thermal Imager - real or fake?

[Short Circuit]

OK, why do they need WiFi to show IR images?  I would think with everyone calling them scammers, they would put the WiFi and the app aside for now and just demonstrate that the camera works.

Yes they do. How else does an image get from their PCB to the phone? If the link is WiFi, and WiFi is not working, then no link.

Nonsense, they bother to produce 'simulated' IR footage. Why not put some efford in simulating that wifi link by tranferring images with SD card, or simulating the app display on a PC?
Nobody cares if they can produce some weird app (which looks utterly useless juding from that video). All doubt in this project is about the thermal imaging, not about the bells and whistles...

[firewalker]

I'm guessing the sensor will communicate with the Wi-Fi chip via a know protocol and not something exotic. If my problem was the Wi-Fi I would post something like

"This is the senor, and this is the raw data when pointed to a heat sink at 50 degrees"

Alexander.

[tom66]

Is there a low-latency protocol for transmitting live video over WiFi? Need to maintain constant frame rate and sync with the non-thermal camera.

I don't know why they used WiFi. Don't most phones support acting as a USB host now? Seems like it would be cheaper and more reliable. Maybe a little harder to write an app for, I don't know.

[sleemanj]

My opinion on the latest video YAR - Yet Another Rendering.  They haven't even shown the "app" running on an actual phone, probably was just an animation.

All these guys have to do to quell 90% of the doubters is to turn on the video camera, pick up their prototypes in their own two hands and SHOW THEM.

Doesn't matter if it's not complete, even if it's not working. 

[EEVblog]

OK, why do they need WiFi to show IR images?

They don't, it's a bullshit delay tactic. It seems obvious they don't have any hardware to get a thermal image out of in the first place.

[EEVblog]

All these guys have to do to quell 90% of the doubters is to turn on the video camera, pick up their prototypes in their own two hands and SHOW THEM.

They can't do that because they don't have any hardware.
So they'll continue to give the bullshit excuse that they need to protect their trade secret.

[EEVblog]

Yes they do. How else does an image get from their PCB to the phone? If the link is WiFi, and WiFi is not working, then no link.

There could be many ways, you build those sorts of things into your hardware for development and debugging.
Remember, it used to be a USB interface camera, not WiFi. They only made that switch way after the campaign ended.
Remember, they have already claimed they are getting thermal images out of their supposed camera. Any excuse about not being able to show an image because WiFi is not working is therefore complete bullshit. There is absolutely no doubt about this based on what they have said. So they have either been lying, or are full of crap, there is no third option.

[Keef Wivanef]

Tried and tested formula.
New Green Widget...coming real soon.
Send money.

Now with crowd-funding, social networking and Google we are living in a scammer's paradise.
If a private citizen pulled stunts like these he could be charged with obtaining money by false pretences, fraud etc.
Register a COMPANY and then you have the GREEN LIGHT.

(cartoon of Keef banging head against wall)

[amspire]

There could be many ways, you build those sorts of things into your hardware for development and debugging.
Remember, it used to be a USB interface camera, not WiFi. They only made that switch way after the campaign ended.
Remember, they have already claimed they are getting thermal images out of their supposed camera. Any excuse about not being able to show an image because WiFi is not working is therefore complete bullshit. There is absolutely no doubt about this based on what they have said. So they have either been lying, or are full of crap, there is no third option.

Don't get me wrong - I think at best, the project has huge problems. I gather no-one know the status of the sensor and lens.

The camera will be a WiFi access point, so I gather that means to use the camera at home, you have to disconnect from your home WiFi and connect to the camera. They you forget you are not connected to your WiFi you start and start watching a whole lot of movies on your phone all charged to your phone account's data plan.

If the WiFi has no security, perhaps it is possible to spoof the camera's WiFi and get someones phone to connect to another rogue access point.

If USB is too hard to get working across all cameras, Bluetooth would have been the right solution even if it meant reducing the video to 1 frame/sec which would be just as useful, but not as funky. Bit late now, since they have promised 30 fps.

I didn't see any reference to getting images out of the camera sensor yet - I did see a vague statement like this:

" Image quality is of utmost importance to us and much of that work is on the hardware side."
"Thermal imaging is tricky, and we want to get it right for every camera we build, so we keep testing and redesigning and perfecting, I think that we’re nearly there though."
" a working camera will be unveiled in early May, and you will see working videos of our camera well in advance of that."   Hmmmmm   

All the posted IR images and videos was a bit misleading - I am sure some people would have thought they were taken with the Mu Optics camera instead of something that probably costs over 20 times more.

I would definitely agree that the camera testing and the phone connectivity testing should have been in prototyping before the fund raising. Now they are committed to delivering about 1800 cameras and they are still trying to work out how to do it. I think they are in trouble. I gather there has been no tooling for the case yet. With the changes in chipsets, they would have to design, order and build new test/prototyping boards before the next round of tests.

Probably an excellent case study in what not to do if you ever want to start a fund raising project with a chance of a successful outcome.

I definitely wont do it this way when I start my IC Printer crowd-sourcing project. I will have the Android and iPhone apps with the big "PRINT IC" button completely working before I start the funding campaign. 

[AlfBaz]

I definitely wont do it this way when I start my IC Printer crowd-sourcing project. I will have the Android and iPhone apps with the big "PRINT IC" button completely working before I start the funding campaign. 

Will it come with built in IC templates??!!?? 

[amspire]

I definitely wont do it this way when I start my IC Printer crowd-sourcing project. I will have the Android and iPhone apps with the big "PRINT IC" button completely working before I start the funding campaign. 

Will it come with built in IC templates??!!?? 

Better than that - it will come with built in ICs!!!!! The design is continually being refined to give you the best possible quality.

Hey, this is easy!

[EEVblog]

" Image quality is of utmost importance to us and much of that work is on the hardware side."
"Thermal imaging is tricky, and we want to get it right for every camera we build, so we keep testing and redesigning and perfecting, I think that we’re nearly there though."
" a working camera will be unveiled in early May, and you will see working videos of our camera well in advance of that."   Hmmmmm   

"nearly there" with "testing and redesigning and perfecting" means you MUST have at least a thermal image. Otherwise you are just full of crap!
Also, they said this:

Abe is in the other room right now tweaking bias voltages and the like to get the best picture possible in the field. Turns out it’s just as hard to make the picture look great as it is just to get a picture in the first place.

If that's true then that means they must have a thermal image. Why don't they show it?

All the posted IR images and videos was a bit misleading - I am sure some people would have thought they were taken with the Mu Optics camera instead of something that probably costs over 20 times more.

I would definitely agree that the camera testing and the phone connectivity testing should have been in prototyping before the fund raising. Now they are committed to delivering about 1800 cameras and they are still trying to work out how to do it. I think they are in trouble. I gather there has been no tooling for the case yet. With the changes in chipsets, they would have to design, order and build new test/prototyping boards before the next round of tests.

They said this 22 days ago:

We’ve got a new case about to be machined (hopefully with proper and final mounts and clips).

That means they have an old case too. They have shown neither.

They also said this:

The next PCB we order should pretty much be the one that goes into the first cameras we ship…baring any big changes

That means they have an old PCB too. They have shown neither.

There are just so many things that show they are talking rubbish. It all points toward no hardware existing.

[Rasz]

Now they are committed to delivering about 1800 cameras

No they arent, this is indiegogo, they are commited to TRY and deliver "something" maybe
You fund a project, not pay for goods.
You funded this guys new car, holidays, maybe a debt he had to repay from his other business (which happens to be 'using special effects to make FAKE VIDEOS LOOK REAL"). 

Also, they said this:

Abe is in the other room right now tweaking bias voltages and the like to get the best picture possible in the field. Turns out it’s just as hard to make the picture look great as it is just to get a picture in the first place.

That means they bloody well have a thermal image, otherwise they are completely lying.

This is an old "goods are in the other room" con trick.

http://youtu.be/fwIW5hnYRi8?t=2m34s

[CanadianAvenger]

All the posted IR images and videos was a bit misleading - I am sure some people would have thought they were taken with the Mu Optics camera instead of something that probably costs over 20 times more.

Indeed, and it is why I initially backed, as by all appearances they had a working prototype. I was pretty pissed after when it came to light that they were taken by a commercial camera, After close of the campaign and identified by a backer and then finally confirmed by Mu, which leads me to believe they had no intent on telling their backers that. I'm actually leaning towards outright fraud now. I'm sure all these backers pulling out will become the ultimate excuse for the projects ultimate failure.

[amspire]

Also, they said this:

Abe is in the other room right now tweaking bias voltages and the like to get the best picture possible in the field. Turns out it’s just as hard to make the picture look great as it is just to get a picture in the first place.

If that's true then that means they must have a thermal image. Why don't they show it?

The wording is very careful - they didn't actually say that Abe has a thermal image from the IR sensor. He might be looking at the sensor output on a scope and adjusting the bias voltages for the best apparent dynamic range of the signals seen on the scope. Sometimes you can see more on a scope output then an actual picture image output which is why waveform monitors were such a big thing back in the analog TV days.

If that is what Abe was doing, then if they tried to post photos, it would scare people off rather then reassure them. To a non-technical person, it might go down as meaningless gobblygook. To a technical person, it would seem that the testing is at an extremely early stage - like the very first stage - of development. It is the sort of test you would do on a breadboard within a few days of getting the first sensor. Definitely not at a "2 months to delivery" stage. What that statement says to me is that the sensor probably needs a dynamically variable bias control to get the best image and it sounds like the planned design is trying to use a fixed bias voltage for simplicity. With a dynamic bias control, you could get the maximum possible dynamic range from the sensor without the need for Abe's "best apparent dynamic range" adjustments.

They have always said it would be a simple design so perhaps it is unacceptably simple.

Reading the Mu Optics page on the Indiegogo site, I have noticed that all the wording is extremely careful and guarded.  It is possible that they have never actually lied so far, but from the reaction of other people here, it seems there is a definite gap between what they actually said, and what many people think they were saying. Words are very important, and when statements are expressed on odd ways, it is a good idea to ask why. If you go through the comments, there are many curiously expressed statements and I have always assumed they were the statements that put the best possible spin on the situation.

My own assessment is that the Mu Optics founders are not electronic designers so they needed the funding money to be able to pay for designers - electronic, software and mechanical. I think this from the fact it is Abe testing the sensor, and not John himself. If it were my project, I would be testing the sensor - the whole project depends on it and I would want to know as much as possible about it. They perhaps did not understand the difficulties in producing a device that may appear to look simple. They did everything possible in the campaign wording to successfully raise as much money as possible from what appears to be a raw concept probably backed up with some discussion of pricing with suppliers.

The idea itself has genuine potential, but I have not yet seen anything to convince me this is the right group of people to make the idea a reality. I would like to discover that they are really competent, but have had to keep a very tight lid on their development work for unknown reasons. Based on what I have seen so far, I wouldn't risk my own money on this campaign.

[andersm]

I don't know why they used WiFi. Don't most phones support acting as a USB host now? Seems like it would be cheaper and more reliable. Maybe a little harder to write an app for, I don't know.

Disregarding the speculation about fraud, to connect a gadget to an iOS device you have to be part of Apple's MFi licensing program, plus the newest models have a different connector to the older ones. I don't know if there are any limitations in Android devices, but at least in theory a wireless connection would be simpler and more universal.

[Bored@Work]

I don't know if there are any limitations in Android devices,

Device manufacturers do whatever pleases them. Often they don't care much about USB. USB support in the Android API is anyhow not too great. Officially you have some Android special (non USB-standard), but documented, accessory support, and rudimentary raw host mode support in Android.  Almost all USB support from the underlying Linux is not available in Android or requires fiddling to make it work. It is a mess.

TCP/IP (over WiFi, mobile broadband or sometimes Ethernet) and Bluetooth are better choices if you need to connect something to Android.

[CanadianAvenger]

The wording is very careful - they didn't actually say that Abe has a thermal image from the IR sensor. He might be looking at the sensor output on a scope and adjusting the bias voltages for the best apparent dynamic range of the signals seen on the scope. Sometimes you can see more on a scope output then an actual picture image output which is why waveform monitors were such a big thing back in the analog TV days.

True, but then they did also post this in an update about a month ago:
"Firstly, I’d like to assuage any fears that we are only now working on WiFi. We have been developing that feature for some time now, and it has been working great. We believe wireless connectivity is a huge step up from USB. The software team has video coming in over the air and it is working flawlessly."

So if they have had video working flawlessly for the past month, why haven't we seen any of it? Answer: They don't have anything, beyond an idea that they can't get to work.

All they need to do is post some images taken with the sensor they are using, even if from an eval board for the sensor they are using with standard optics. For the money they raised, you'd think they could at least have done that!  But now they are too far down the road of their deception, and will need to do substantially more to assuage any doubts now. The backers should be demanding images of an actual PCB with some markings that show it is the Mu optics camera... this can be done without revealing any of the "secret sauce" they are so paranoid of revealing.

[BravoV]

After reading through the progress there, and also replies/soundings here in this thread, why I have the feeling that they are really aware that every updates, or words that are coming out through their mouth are very thoroughly crafted. Damn sure I'm not alone here.

Its like a really hard work and great effort for the sake to maintain the period for the benefit of doubt as long as possible until it last ?   

[amspire]

True, but then they did also post this in an update about a month ago:
"Firstly, I’d like to assuage any fears that we are only now working on WiFi. We have been developing that feature for some time now, and it has been working great. We believe wireless connectivity is a huge step up from USB. The software team has video coming in over the air and it is working flawlessly."

So if they have had video working flawlessly for the past month, why haven't we seen any of it? Answer: They don't have anything, beyond an idea that they can't get to work.

This is another carefully worded statement. They said "Video", not "IR Video from the thermal sensor". If they only say "Video", I assume they only mean "Video" and absolutely nothing more.

In another statement, they said they had to discount Bluetooth as is could not sustain 30fps video, so it makes a lot of sense they test to see if WiFi can sustain 30fps video. That video may have been a simple test pattern generated by code in the micro - it could be done very simply. It is much more like they were just testing WiFi in general without using any of their hardware. They may have been sending a video stream from a notebook PC to their App running in the android development system on another PC via WiFi purely to test if the App could process a 30fps video stream arriving over WiFi without dropped frames. That might be how they got those demo videos into their demo App that they posted.

Back to the words "The software team has video coming in over the air and it is working flawlessly". That is all they are saying. If they had an IR camera talking to a phone, they would have used different words for sure.

The words "Software Team" means they probably have two programmers. One developing the Apps. One developing the embedded code in the camera.

On a positive note, it sounds like they are trying to develop something. It doesn't sound like there are two people sitting on a beach in the Bahamas with a big bag of money. The camera is a really good idea, and the best thing for contributors would be if the cameras are not rushed out in two months, but instead if Mu Optics gets proper funding after they get a working prototype (or even sell the startup to a rich company) so that you have a chance of getting a more mature design in perhaps 12 months. $300,000 may sound a lot, but it is not when the money is meant to pay for the parts and manufacture of 1800 cameras to the contributors, plus wages for a team of probably at least 6 people for at least 6 months, plus case designing, prototyping and tooling, plus radiation testing, plus equipment, rent, electricity and so on. The idea is good enough to earn more then the $300,000 campaign funding, and I thing the founders will be hoping for more then $300,000 from the project.

[tom66]

I'd like to hope this isn't a scam. (I haven't invested, I'm just making a point.)

My reasons -- They've got the money, there's little or no chance of refunds happening. And they've got what appears to be staff, in an office, somewhere, although it does look more like a converted house. (And a small one...)

They've also developed an app which although presently fake does demonstrate some functionality.

Either way they've gone to a lot of effort to either fake it (which seems unnecessary IMO) or they've realised how much work there truly is behind getting something like this to market and they're stalling as they try to get it out there.

[CanadianAvenger]

My take is that it isn't a scam, but it is fraud -- as there was serious misrepresentation as to the state of affairs of the project. While they may have been careful in their wording by not explicitly saying they had something, the natural inference is that they did because of what they were showing and saying. They needed to be clear that what they were showing was not from their prototype to avoid misrepresenting their position.

My guess is they are now feverishly trying to get things to a state that they can show. Assuming they were starting from scratch at the conclusion of the campaign,  if they were good, they could have spun a first working prototype in about 4 weeks from the go point, which means they could have shown a PCB by now. My guess is that they don't have the level of expertise required to turn the hardware so fast [and possibly not even enough to ever get a working prototype], so it will be significantly longer. The fact they are still figuring out some key components would suggest that no PCB has even been spun yet. [though they indicate that one has by alluding to the "next PCB" in one of their updates] There is deliberate misdirection and omission in their communications, which all leads to my conclusion of fraud by misrepresentation.

[Rasz]

Either way they've gone to a lot of effort to either fake it

The biggest effort so far was using special effects in form of clever image processing algorithm to map RGB values into false color, in effect creating pseudo thermal image with zero value other than misleading.
Interesting facts about the video. Information about picture being simulated and not from real thermal camera is edited over real audio. Clip was quickly taken down and reuploaded - just enough time to read a comment congratulating them on amazing breakthrough in thermal imaging technology that let them grab thermal reading THROUGH A GLASS WINDOW, take it down, and redub the audio.

(which seems unnecessary IMO)

Good convincer is everything in long cons.

[mikeselectricstuff]

If it had been a scam from the start, they could have been much, much more convincing than they have been, and would have cashed out and run by now. There would be no reason to have carried on this long.
My guess is they just didn't do enough research into the problem, steamed ahead with an "It'll be alright" sort of attitude from a (possibly self-delusiuonal) leader with a salesman type attitude, assuming cheap thermal imaging was something they could do with off-the-shelf parts and a moderate budget, and by the time they realised what they were into, it was too late so the bullshit ass-covering started to avoid losing face.
 A possible alternate scenario is they were over-promised something by a subcontractor (or "University genius" type who had an idea but no clue about manufacturing), let down, and are now drifting further and further up shit creek with no means of propulsion.

The one thing that has made this stand out for miles as vapourware is that at no stage in all of this have they ever mentioned anything whatsoever about the core technology - everything, from the initial pitch to the subsequent bullshit-filled updates has been all about entirely peripheral issues  like housings,  connectivity, getting it to retail etc. which are all much, much easier and readily solvable than the core functionality, of which they have demonstrated absolutely ZERO evidence.
 
They have nothing, will never have anything, and the only questions are how long it will be before it all implodes and the top people either do a runner or get prosecuted, and whether there will be any money left to compensate the backers.

[tom66]

In terms of sensors they may be using devices like these
http://www.ulis-ir.com/index.php?infrared-detector=25--m-160x120hv

These devices are promising. My understanding is FLIR's patents expired a few months ago and that's why were seeing a rush of new sensors appearing. I'm mighty tempted to ask for sample prices of these, but I expect it will be in at minimum the $100+ range. Perhaps they've realised that cheap thermal sensors... aren't.