I read-up on the machines purpose. It was used in cancer research, where one could take up to 48 cell samples. While keeping them alive up to 24h within the sample well plate. During this each cell 'channel' can be treated by different chemicals. In the same different fluorescent dyes can be dispensed per cell, different other reagents for instance to test cells viability / metabolism. I think the easiest way to think of it, is like a human bio-reactor where different cancer treatments can be test in parallel and rapidly.
Those FLI cameras use pretty standardised control software, looks to be a custom version of their Hyperion range, would be a killer astrophotography imager on the the right telescope.
Was my first thought as well, that huge sensor would be awesome in an astro setup. Well, maybe - don't know the exact specs of it, but presumably it's "pretty good".
I used to work for a much smaller similar company. Here's the catch 22, these are low volume high cost devices without the designer wankery. Customers come to the manufacturers lab and say, "This is great, it does exactly what I want. But it's expensive and doesn't look that nice, what else can you do to make it look better?". I paraphrase but that's basically what happens. So then you go down the design rabbit hole and end up with this monstrosity, and you charge the customer twice as much but for some reason they now want to pay it. It's broken. Also Fluidigm are probably the most guilty of doing this.
Well said. In a world where all of these tech startups survive by wowing VCs who pray to Steve Jobs, it probably pays just as much to have a pretty lab as it does to have a lab that actually does something. Pretty funny from a scientific/engineering perspective. But I guess you do what you gotta do to survive, including burning piles of cash for milled titanium bezels.
Note to design group: Budget unlimited, I want it totally bespoke, make sure it looks good in the operators living room, and make sure you use the CNC division of our company a lot.
As someone not directly involved but very familiar with the industry of super-expensive super-low-volume machines: The looks do matter. In the end, these things get sold at tradeshows. Having your unit stand out on a trade show makes a big difference. On top of that, it is a way to differentiate from others. By not being 'boring grey box', you avoid being forgotten amongst the sea of grey boxes. Then there is the psychological effect. Sure, that machining might add 5k to the BOM of the machine, but after all, that is just 1% of the list price. And the users who might trial the machine do subconsciously (even if people claim the don't care - research shows they do) associate it with a fancier, better product, even if the specs are almost identical to another machine. And then if your device is in the same price bracket as another device with similar specs, you might win the contract a bit more often than if you saved a few k on machining. (that said, some of the machining here is a tad excessive, like the display - could make that a straight back with much fewer machining ops that would still look amazing, but I digress). It also helps you have a clear, recognizable style. Why is all the expensive keysight equipment black, rohde&schwarz blue-and-white, and Anritsu Green-and-white? Because that kind of corporate identity matters to your end customers. Even if all of this just sells you 10 or 20 extra machines, on a machine you might sell a few hundred of, that makes a difference. On top of that, you might get away with a few percent higher sales price, which easily offsets any additional cost of the machining. And as a final fact: A lot of these product lines start with startups. Those startups have over-eager (young, fresh grad) designers and engineers who are just excited and lose track of the bigger picture.
Yes will bet the ones in those cases are original designs, the later ones in less fancy are from later on, when the company was bought out by one of the larger ones, and only keep the old machines in the line up because they have either lots of the parts in stock, and the income from consumables is good. Later on they will be discontinued as parts run out, and then the boring generic line will be folded into the parent completely, leaving no trace it ever existed.
You're not wrong! But I have to admit that if I had to select what I buy for my company, I would already be put off by the power button not being a physical, reliable clonky switch. I would also not like to spend extra on some useless design, which I KNOW is only there to lure me towards that model. I hate being lured and advertised to with quite some passion. I want specs, certificates, TCO metrics and a fair price. I already know that I won't be getting the latter with products like that. Unless it absolutely excels in its actual task, and/or is EXACTLY the package I need, with alternatives with a broader use costing noticably more. Otherwise stuff like that would absolutely warn an shoo me away. Anyone responsible for purchases at those prices that leans towards wanky bling is definitely in the wrong position, if you ask me. Edit: Case in point, as I've continued watching now: I would have NEVER expected this thing to be so very well and solidly engineered and built, based on its "lets just make some suits happy" appearance.
@@fonkbadonk5370 I hear you! Definitely think this unit pushes it maybe a tad too far, and get that it can turn people off too! Regarding the power switch: Probably done because you want the unit to power-down in a predictable state. You can't have all the valves sit in some unknown condition before moving, etc...
@@JorenVaes It doesn't have to be an actually power-cutting switch. Just a physical one that gives off a nice definitive "clack" when I push it. Like actual metal contacts making contact. It may fall back on its own and just signal a shutdown, but I want to know it won't easily break, and can be operated with gloves or otherwise "compromised" fingers (which is an accessibility concern on top).
That image sensor is massive, hopefully it's a decent modern one. The Windows UI looks really neat too, seems they used all the modern UI tech which was pretty much dropped right away. It looks very much like the Windows Phone style.
This reminds me of why I usually loathe working with industrial designers, especially the ones on the artsy end of the skill spectrum. Grumble, grumble.
Did they hire the same designer as Juicero? While I'd usually refer to the light effect as "breathing", "throbbing" seems particularly apt for the amount of wankery going on with this device.
For something with a lot of impressive engineering, there is also a lot of absolutely atrocious engineering. I have a manufacturing background, and those giant monolithic aluminum components are absolutely insane. You're probably looking a close to $1500 in aluminum billets for those two parts you showed at the end, and then throwing away most of it with $4000 worth of machining. It wouldn't have been as elegant, but breaking those components up into several smaller pieces that were bolted or welded together would have seen considerable cost and manufacturing time savings. Couple that with the side panels milled from aluminum sheet, and it's just craziness. "Spare no expense" doesn't mean "design this purposely to be unnecessarily expensive to build". The whole thing screams "biotech engineering intern designed". Bunch of very intelligent, creative college kids, without real-world engineering experience, being told they have no budget constraints. I just cannot see a seasoned mechanical engineer making those pieces out of solid aluminum billets like that, even with an unlimited budget. It's just so wasteful.
well, they sell it 400K, so they are not a few thousandth. This kind of "engineering" is actually quite common in "tech" company making insane markups.
Although I agree with you from a perspective that you shouldn't be wasteful, I don't think doing it in a different way is necessarily cheaper (or better). A good example what else might tie in is time (engineering hours). If there is extra engineering time involved into creating parts that bolt together nicely, it could costs the company money as well. Engineers are expensive, and sometimes the seemingly more expensive route is actually cheaper in terms of man power. This is often the case for low volume products, where R&D/design costs make up a much larger portion of the final unit cost compared products sold at volume, where the engineering cost can be spread over many more units. That is just considering the cost of the engineers, not even valuing other key drivers such as the lower time to market (the sooner you can get your product to the customer the better). I work in research, and there too we often go for the 'simple' solution, even though it might be expensive on the face of it, just because spending a few days more on it to drive down the cost, actually costs more in the complete scope of things. Tl;dr there are more key-drivers involved in making design choices, and something which seemingly is a more expensive option on the surface might actually turn out cheaper if you factor those in as well.
Probably not relevant here, but after working in a hospital in a none clinical role I have seen the criminal exploitation of patients by pharmaceutical and medical device suppliers. It is staggeringly. I would include consultants as well. Healthcare is an aggressive industry targeting people who are frightened. Money is there to be extracted and they are experts at that. "We're all about the patient" is a phrase often said
That manifold is beautiful. Get it to Lookmumnocomputer to make a miniature whistling pipe organ from it! Great vid Mike. Looking forward to the camera exploration
Definitely some very interesting stuff going on inside hat machine. 400k is something! I'm kinda envious about how you're able to lay your hands on high-end scientific and medical gear, but now that I saw how Rinoa got a high-performance liquid chromatograph that was no longer used and just sitting there in a basement in some lab, everything is possible. I love the fluidics part of it, it's extremely complex. A thing of beauty and a joy for ever... maybe not that for ever now that you discombobulated it. Definitely some wonderful engineering went into it. Agreed on using a PCB as a backplane for solenoid valves. It'd be so much more elegant his way. I can imagine Dave Jones aesthetically wanking over the die-cast aluminium parts, haha! 16:57 who even does check flags using case-sensitive string comparison like that? Totally unpythonic.
„aesthetic wankery“ 😏 Oh well, in medtech there is nearly no limit to the cost of the outer surfaces, especially if they have to comply with GLP regulations. I‘m slightly surprised at the uneven surface outside as it doesn’t seem to be beneficial for cleaning. Have yet to watch the whole series but I assume they‘re using the massive case to keep the inner temperature within a defined range compared to the temps in the lab it would be installed in. 400k is not a surprising price to me at all, yes that outer finish is for the fuzzy feeling in this case. The microfluid consumables are single use and they will be the main cash cow next to the sensor. Fascinating stuff, don‘t get to see this every day.
These devices are engineered and validated for a specific purpose and in the end that purpose is what will define the requirements. The cost isn’t the main deciding factor, the question is if the device and the manufacturer can fulfil the requirements posed by the customer. Those devices must look neat to the cleaning crew, like the synchronous blinking BMWs.
Yeah I mean at its core is a bunch of serious engineering. My guess is they got that all working, then the wigs said "Make it pretty. Also you have no budget constraints".
Thanks Mike! Good recommendations at the end as well, wonder if the makers could share some thoughts on that. Not the case for this one but sometimes companies have to make a stupid design choice such that the integration steps can be verified.
On one hand, perhaps some of the design is a bit wanky and overdone. But it looks like the UI has had at least a little bit of thought put into making it nice to use, and look decent. That alone puts it leagues above most specialized equipment where the UIs are generally ugly and unnecessarily obtuse (and still running on Windows XP)
The heaters were probably made by Heatron in Leavenworth, KS which is up in the Kansas City area. They are one of the largest producers of thick film heaters.
Just from seeing it before it reminds me of Theranos' Edison prototype. That already was very much working that "Apple" vein. With that, this or a Juicero we're seeing what happens when your design and product costs are unconstrained.
The code for that script... spills wonders on its creator. They really spent the money on looks, but the parts that does the actual job, seems to have been second thought.
never underestimate the ability of C-suite types and potential investors to be impressed by rows of shiny equipment that they have no idea what it does. the shinier the better, obviously. squirrel!
16:58 That script is python 2. If you have write access to it (for example if you pull the drive and put it in another pc) you could rewrite a section of it pretty easily and get it to dump the passwords or even just execute the maintenance mode directly.
@@randacnam7321 Yeah the only similarity is the insane machining/material resources used for prettiness. Once you get inside there is nothing to compare to the great bag squisher. Seriously awesome engineering inside.
Small valves rely on contact pressure, they probably replace the wiring section to a valve with the valve, so that they only are used once. That way the initial wipe keeps the connection gas tight, and it will not vibrate loose. Yes pneumatics dept and electronics did not talk much to each other, or they had a massive delay on laying out that pneumatic logic board, so did not actually know until late on in the design as to where they were going, how many they would need, and an exact design, as they probably were iterating through designs, trying to get it to both be manufacturable with some sort of yield (likely they had a 50% loss in testing, which is fine if you are charging $400k, make around a dozen a year, and the fluidic assembly takes 2 days to bond together, and another day to pass or fail test, so you always have a stock for testing) that is acceptable and not going to delay any deliveries.
That optical block is probably made from Invar, as it has almost zero TC, so the focus does nor change with temperature, and also will nor change with the top being cooled and the base being heated.
There are similar instruments that actually use 300mm-wafer image sensors, a single sensor per wafer. Designed by companies such as Caeleste in Belgium.
The attention to detail at 24:40 is incoherent with the "wankery" at the beginning.. almost as if two separate designers were at work, one for the guts, and another for the lipstick.
Spotted a Qioptiq logo there, makers of very high end commercial and military-spec optical stuff. I guess all the fancy visual LED and metal machining is for when the venture capitalists get a visit to the latest company they spun out. They can see how their money was spent!
No wonder it was 400.000 US$, just imagine how many people with specialised knowledge were engaged in developing, designing and building these things and its not like they sold in millions.
Thanks for the video, it's really interesting to see these teardowns of very specialised technical equipment. Probably not cheap (for you) to buy either.
Ah! Reminds me of the failed attempt from eddify, in a different yet also niche field (NDT) to do something similar. Plus, they even added subscriptions to use the hardward, and have it bricked if you dont renew :)
I think it appropriate that the external case and monitor should somewhat be aligned with the internals. Would it make sense to have the external body/interface be minimal (read:cheap/standard dull) when it is
Company to their engineers: We need a medical machine but it can not be to expensive but we give you all free hands so you can do whatever you want to! The Engineers: Never say that to an engineer, this is what would happen..
I love engineering undertaken without the control of the accounts department. It looks like they could halve the bill of materials without effecting the functionality. Good score Mike. Thanks.
There exist many entities that thrive on generous government grants by finding ways to spend ever more money on such things. In such cases the high price becomes a welcome feature.
@@NeungView You know what... I must have been hallucinating. None of the entities in biotech, healthcare, pharma, academia and certainly not in the defence industry who receive government grants would _ever_ seek novel ways to spend large sums of money in order to justify increases in future funding. My bad!
Definitely resin printing behind those translucent fluid manifolds. Health insurance markups & a quantity of 1 would make it cost 1/6 of a house. Guess we'll have to buy $400k of mandatory health insurance & the net investment income tax won't be tracking inflation.
I can't disagree with most of these comments, especially the "Apple-level aesthetic wankery". OTOH, If your health-care team (and mine) are relying on machines like this to diagnose what is wrong with you, or to analyze what kind of therapy is most effective to make you better or keep you alive, I am somewhat reassured at the level of insane build-quality and intricate custom manifolds, etc. and automation of what was probably a rather fiddly series of laboratory tests. This illustrates how much mechanical/electrical/chemical/billogical/optical automation is taking over things that literally affect our life, health, and safety. And when A.I. starts reading our medical charts and talking to machines like this, we are handing over the welfare of human society to the machines. Let's try to not interface things like this to "Skynet" (or SKYNET). 😲
There is a supplier for Electric and Phneumatic PCB near Hamburg. Saw their stuff at the SPS show in Nuremberg in 2016. So it is available, but the number of layers in this manifold is extreme. Although as you say a simple thin multilayer, or even a flexible foil PCB could have done a better job than the extremely expensive wire loom. But looking at the overall concept of this machine, there was little mind given to optimization. At a price of 400k there was surely a fair amout of profit in this. So, inguessbthete is a lot of scope to bring that kind of device to a much lower price point by a less lavish implemebtation an the application of a little judicious production optimization. I remember seeing well plates at a Medica show in Duesseldorf in the early naughties, but I never saw a machine that drove them. How did you find sth like that in the first place?
I wonder how they design those microfluidic stacks. Perhaps they just use regular PCB layout software with special footprints, extract the different layers from the gerbers, edit some stuff manually and use those for machining the plates before bonding?
what? lol, why would you use altium, it would make zero sense, you can be sure there is specialized CAD sotware to do that, if not they just used industry standards like Katia or solid edge.
@@lo2740 I never mentioned altium specifically. And why exactly would you not use PCB layout software, why does it not make sense? The whole stack is just a 2.5D problem, no need for full 3D modelling. Essentially these stacks are also just traces on substrate. Instead of copper it is flow channels (lack of material) and instead of substrate you have poly carbonate. The output of PCB software can be chosen as vector files, and those are perfectly fine inputs for CNC routers. If you can CNC the plates, and stack them together, you have your microfluidic stack. You tell me why that 'makes zero sense'. I'm sure there's extensions for Solidworks etc, but it doesnt make it inviable or dumb to think of other ways to do that same task, perhaps even more efficient.
You need the threaded holes and other features which aren't used in PCB CAD. So solidworks, katia, etc. as mentioned would be used. You can also use them to do fluid simulations, calculate tolerances, part weight, and more.
Heck of an image sensor. It would be interesting to see what that is capable of. Shame there's probably not a lot that can be done with it otherwise without a whole bunch of very custom work. But you can definitely see where all that ungodly weight came from. All that metalwork for the optics and whatnot was pretty wild to see. Though I can safely say that the external case just looked stupid to me. Probably an absolute crapton of machining to make it look like a grey block cut out from a bigger chunk of something by someone using a one inch drill bit. Strange.
Probably easy enough to interface, USB, though likely a lot higher resolution than what they record. Probably used because they have a higher end version that uses it, so use the larger sensor, and crop the image. Makes alignment easier though.
Wow they really gilded the lily on that one, for no good reason. Seems like a mechanical engineer had a wet dream or something on this thing. The expense paid for some of it's features is really comical (like how they did all that just for a fancy door movement).
Biotech is trending towards appleization. Hgih end (millions usd) equipment with minimalistic interfaces, mechanics so complex you need a specialized technician for anything, and propietariy softwares that are often hard to obtain. Difficult to do science in poor countries
Unbelievable. I guess some biotech group couldn't swing buying a company Rolls and had to settle for a fancy bit of test equipment they didn't really need.
I read-up on the machines purpose. It was used in cancer research, where one could take up to 48 cell samples. While keeping them alive up to 24h within the sample well plate. During this each cell 'channel' can be treated by different chemicals. In the same different fluorescent dyes can be dispensed per cell, different other reagents for instance to test cells viability / metabolism.
I think the easiest way to think of it, is like a human bio-reactor where different cancer treatments can be test in parallel and rapidly.
Watching someone reverse engineer with so much knowledge is the most satisfying thing. Thanks as always Mike!
Those FLI cameras use pretty standardised control software, looks to be a custom version of their Hyperion range, would be a killer astrophotography imager on the the right telescope.
Was my first thought as well, that huge sensor would be awesome in an astro setup. Well, maybe - don't know the exact specs of it, but presumably it's "pretty good".
I used to work for a much smaller similar company. Here's the catch 22, these are low volume high cost devices without the designer wankery. Customers come to the manufacturers lab and say, "This is great, it does exactly what I want. But it's expensive and doesn't look that nice, what else can you do to make it look better?". I paraphrase but that's basically what happens. So then you go down the design rabbit hole and end up with this monstrosity, and you charge the customer twice as much but for some reason they now want to pay it. It's broken. Also Fluidigm are probably the most guilty of doing this.
Well said. In a world where all of these tech startups survive by wowing VCs who pray to Steve Jobs, it probably pays just as much to have a pretty lab as it does to have a lab that actually does something. Pretty funny from a scientific/engineering perspective. But I guess you do what you gotta do to survive, including burning piles of cash for milled titanium bezels.
Jesus £120 for all that tech, you got a right bargain there mike 😂
That slide out draw mechanisim with drop down door was pretty impressive, so was the size of that CCD. The pneauumatic manifold was art!
Note to design group:
Budget unlimited, I want it totally bespoke, make sure it looks good in the operators living room, and make sure you use the CNC division of our company a lot.
As someone not directly involved but very familiar with the industry of super-expensive super-low-volume machines: The looks do matter. In the end, these things get sold at tradeshows. Having your unit stand out on a trade show makes a big difference. On top of that, it is a way to differentiate from others. By not being 'boring grey box', you avoid being forgotten amongst the sea of grey boxes.
Then there is the psychological effect. Sure, that machining might add 5k to the BOM of the machine, but after all, that is just 1% of the list price. And the users who might trial the machine do subconsciously (even if people claim the don't care - research shows they do) associate it with a fancier, better product, even if the specs are almost identical to another machine. And then if your device is in the same price bracket as another device with similar specs, you might win the contract a bit more often than if you saved a few k on machining. (that said, some of the machining here is a tad excessive, like the display - could make that a straight back with much fewer machining ops that would still look amazing, but I digress).
It also helps you have a clear, recognizable style. Why is all the expensive keysight equipment black, rohde&schwarz blue-and-white, and Anritsu Green-and-white? Because that kind of corporate identity matters to your end customers.
Even if all of this just sells you 10 or 20 extra machines, on a machine you might sell a few hundred of, that makes a difference. On top of that, you might get away with a few percent higher sales price, which easily offsets any additional cost of the machining.
And as a final fact: A lot of these product lines start with startups. Those startups have over-eager (young, fresh grad) designers and engineers who are just excited and lose track of the bigger picture.
Yes will bet the ones in those cases are original designs, the later ones in less fancy are from later on, when the company was bought out by one of the larger ones, and only keep the old machines in the line up because they have either lots of the parts in stock, and the income from consumables is good. Later on they will be discontinued as parts run out, and then the boring generic line will be folded into the parent completely, leaving no trace it ever existed.
You're not wrong! But I have to admit that if I had to select what I buy for my company, I would already be put off by the power button not being a physical, reliable clonky switch. I would also not like to spend extra on some useless design, which I KNOW is only there to lure me towards that model. I hate being lured and advertised to with quite some passion.
I want specs, certificates, TCO metrics and a fair price. I already know that I won't be getting the latter with products like that. Unless it absolutely excels in its actual task, and/or is EXACTLY the package I need, with alternatives with a broader use costing noticably more. Otherwise stuff like that would absolutely warn an shoo me away.
Anyone responsible for purchases at those prices that leans towards wanky bling is definitely in the wrong position, if you ask me.
Edit: Case in point, as I've continued watching now: I would have NEVER expected this thing to be so very well and solidly engineered and built, based on its "lets just make some suits happy" appearance.
@@fonkbadonk5370 The touch power button is there to indicate that the entire GUI of the machine is tactile.
@@fonkbadonk5370 I hear you! Definitely think this unit pushes it maybe a tad too far, and get that it can turn people off too!
Regarding the power switch: Probably done because you want the unit to power-down in a predictable state. You can't have all the valves sit in some unknown condition before moving, etc...
@@JorenVaes It doesn't have to be an actually power-cutting switch. Just a physical one that gives off a nice definitive "clack" when I push it. Like actual metal contacts making contact. It may fall back on its own and just signal a shutdown, but I want to know it won't easily break, and can be operated with gloves or otherwise "compromised" fingers (which is an accessibility concern on top).
That image sensor is massive, hopefully it's a decent modern one. The Windows UI looks really neat too, seems they used all the modern UI tech which was pretty much dropped right away. It looks very much like the Windows Phone style.
This reminds me of why I usually loathe working with industrial designers, especially the ones on the artsy end of the skill spectrum. Grumble, grumble.
I absolutely love these types of videos!!! You are awsome mike!
Did they hire the same designer as Juicero? While I'd usually refer to the light effect as "breathing", "throbbing" seems particularly apt for the amount of wankery going on with this device.
Haha, I was definitely reminded of the AvE Juicero teardown while watching this!
Juicero definitely came ro mind here 😂
That, IIRC, is the first time I've heard Mike 'WOW' or 'just a magnificent thing'.
For something with a lot of impressive engineering, there is also a lot of absolutely atrocious engineering. I have a manufacturing background, and those giant monolithic aluminum components are absolutely insane. You're probably looking a close to $1500 in aluminum billets for those two parts you showed at the end, and then throwing away most of it with $4000 worth of machining. It wouldn't have been as elegant, but breaking those components up into several smaller pieces that were bolted or welded together would have seen considerable cost and manufacturing time savings. Couple that with the side panels milled from aluminum sheet, and it's just craziness.
"Spare no expense" doesn't mean "design this purposely to be unnecessarily expensive to build". The whole thing screams "biotech engineering intern designed". Bunch of very intelligent, creative college kids, without real-world engineering experience, being told they have no budget constraints.
I just cannot see a seasoned mechanical engineer making those pieces out of solid aluminum billets like that, even with an unlimited budget. It's just so wasteful.
well, they sell it 400K, so they are not a few thousandth. This kind of "engineering" is actually quite common in "tech" company making insane markups.
Although I agree with you from a perspective that you shouldn't be wasteful, I don't think doing it in a different way is necessarily cheaper (or better). A good example what else might tie in is time (engineering hours). If there is extra engineering time involved into creating parts that bolt together nicely, it could costs the company money as well. Engineers are expensive, and sometimes the seemingly more expensive route is actually cheaper in terms of man power. This is often the case for low volume products, where R&D/design costs make up a much larger portion of the final unit cost compared products sold at volume, where the engineering cost can be spread over many more units. That is just considering the cost of the engineers, not even valuing other key drivers such as the lower time to market (the sooner you can get your product to the customer the better).
I work in research, and there too we often go for the 'simple' solution, even though it might be expensive on the face of it, just because spending a few days more on it to drive down the cost, actually costs more in the complete scope of things.
Tl;dr there are more key-drivers involved in making design choices, and something which seemingly is a more expensive option on the surface might actually turn out cheaper if you factor those in as well.
Love your evaluation 🤣 Looking orward to seeing the light source and camera!
Looks very Theranos
I thought this thing is the size of a shoe box.. until the tiny hand came in :D
it was only used when ever the price was mentioned in front of it
"Apple-level aesthetic wankery" really got me! :) Unbelievable enclosure, man.
This guy and thunderfoot are are my TH-cam heros!
Probably not relevant here, but after working in a hospital in a none clinical role I have seen the criminal exploitation of patients by pharmaceutical and medical device suppliers. It is staggeringly. I would include consultants as well. Healthcare is an aggressive industry targeting people who are frightened. Money is there to be extracted and they are experts at that. "We're all about the patient" is a phrase often said
That door definitely should have been brought to the attention of Boeing's design team.
So well said about the "Applezation" of some super high end specific hardware
"Apple level of aesthetic wankery"
and markup.
Juicero Biotech LLC
That manifold is beautiful. Get it to Lookmumnocomputer to make a miniature whistling pipe organ from it!
Great vid Mike. Looking forward to the camera exploration
Equipment like that is made to look like that to impress the "suits" when they come on a tour through the lab....
More specifically, the VCs.
Sounds a bit like the scene from Monty Python, this is the most expensive machine in the hospital 😂
@@ZeroStatic But does it go "BING!!!"?
Turned on once a year when the investors come through, then pushed back under a desk for the rest of the year.
Maybe they picked up the people form Juicero after their juicer failed
The assignment here was definitely a bit more complicated than "squeeze pouch". There's wankery on the outside, but the innards are pretty awesome.
Adapt that lovely big sensor into a large format or medium format camera.
This has to be one of the most impressive things you've torn down!
Epic teardown of an epic piece of machinery!
Great video.. very interesting.. looking forward to part 2
Definitely some very interesting stuff going on inside hat machine. 400k is something!
I'm kinda envious about how you're able to lay your hands on high-end scientific and medical gear, but now that I saw how Rinoa got a high-performance liquid chromatograph that was no longer used and just sitting there in a basement in some lab, everything is possible.
I love the fluidics part of it, it's extremely complex. A thing of beauty and a joy for ever... maybe not that for ever now that you discombobulated it. Definitely some wonderful engineering went into it. Agreed on using a PCB as a backplane for solenoid valves. It'd be so much more elegant his way.
I can imagine Dave Jones aesthetically wanking over the die-cast aluminium parts, haha!
16:57 who even does check flags using case-sensitive string comparison like that? Totally unpythonic.
that last part of your comment, lord. I spat my drink
There's something so apple about designing something completely bespoke only for it to be worse, or simply unnecessary.😮
„aesthetic wankery“ 😏
Oh well, in medtech there is nearly no limit to the cost of the outer surfaces, especially if they have to comply with GLP regulations.
I‘m slightly surprised at the uneven surface outside as it doesn’t seem to be beneficial for cleaning.
Have yet to watch the whole series but I assume they‘re using the massive case to keep the inner temperature within a defined range compared to the temps in the lab it would be installed in.
400k is not a surprising price to me at all, yes that outer finish is for the fuzzy feeling in this case. The microfluid consumables are single use and they will be the main cash cow next to the sensor.
Fascinating stuff, don‘t get to see this every day.
These devices are engineered and validated for a specific purpose and in the end that purpose is what will define the requirements.
The cost isn’t the main deciding factor, the question is if the device and the manufacturer can fulfil the requirements posed by the customer.
Those devices must look neat to the cleaning crew, like the synchronous blinking BMWs.
It all ways looks like the engines just got a no limit credit card and little by little went insane in luxury mode.
My engines always get whatever they want
A bit Apple-y yes, but unlike most mainstream Apple stuff, these guy did put a fair bit of engineering effort in that benefits the customer.
Yeah I mean at its core is a bunch of serious engineering. My guess is they got that all working, then the wigs said "Make it pretty. Also you have no budget constraints".
That manifold!! But agree why they just didn't add a pcb for the valves
Clear cam lens 😮💨
Let's gooooooo! It was nice meeting you on the Hackaday conf in Berlin. Hope this video is about what you where showing there!
Love it when Mike gets a bit pissed at the designers! You’re not wrong 😉
it's always a treat when you upload, thanks for showing us this! so.... the machine costs $29.99+s/h, the tests cost $125,000 each? haha
Thanks Mike! Good recommendations at the end as well, wonder if the makers could share some thoughts on that.
Not the case for this one but sometimes companies have to make a stupid design choice such that the integration steps can be verified.
On one hand, perhaps some of the design is a bit wanky and overdone. But it looks like the UI has had at least a little bit of thought put into making it nice to use, and look decent. That alone puts it leagues above most specialized equipment where the UIs are generally ugly and unnecessarily obtuse (and still running on Windows XP)
a python script, lol.
@@lo2740the event scripts are python, pretty confident the main application & GUI isn't. And there is nothing wrong with using python for that.
The heaters were probably made by Heatron in Leavenworth, KS which is up in the Kansas City area. They are one of the largest producers of thick film heaters.
Makes sense that Canadian companies make good heaters!
@@jon1758Actually they're Swedish; they were acquired by Backer a few years back.
Yup the MFPs are single use consumable parts. One way around the 1000 bucks each is to use a photoresin machine and dxf files to create your own. ❤
Just from seeing it before it reminds me of Theranos' Edison prototype. That already was very much working that "Apple" vein.
With that, this or a Juicero we're seeing what happens when your design and product costs are unconstrained.
The code for that script... spills wonders on its creator. They really spent the money on looks, but the parts that does the actual job, seems to have been second thought.
Not really, what it costs to design that outside look is pretty marginal compared to what it costs to design the internals of this thing.
very common in this kind of equipement, software ends up being crappy python script.
never underestimate the ability of C-suite types and potential investors to be impressed by rows of shiny equipment that they have no idea what it does. the shinier the better, obviously.
squirrel!
16:58 That script is python 2. If you have write access to it (for example if you pull the drive and put it in another pc) you could rewrite a section of it pretty easily and get it to dump the passwords or even just execute the maintenance mode directly.
Funny that they used python2 in 2014, which was already announced eol by then
@@JinguapingiI'd say more like typical tbh. Python2.7.xx had a long run..
Juicero is what comes to mind. The yield is quite similar as well.
But this at least has better results with the machine than squeezing the consumables by hand.
@@randacnam7321 Yeah the only similarity is the insane machining/material resources used for prettiness. Once you get inside there is nothing to compare to the great bag squisher. Seriously awesome engineering inside.
One can only imagine what kind of test substance you accidentally squirted all over the camera lens...
Would be fun to make an anamatronic with all those valves.
World's first microanimatronic! A tiiiiny Chuck E Cheese band.
@@ross302ci That would be rad.
Small valves rely on contact pressure, they probably replace the wiring section to a valve with the valve, so that they only are used once. That way the initial wipe keeps the connection gas tight, and it will not vibrate loose.
Yes pneumatics dept and electronics did not talk much to each other, or they had a massive delay on laying out that pneumatic logic board, so did not actually know until late on in the design as to where they were going, how many they would need, and an exact design, as they probably were iterating through designs, trying to get it to both be manufacturable with some sort of yield (likely they had a 50% loss in testing, which is fine if you are charging $400k, make around a dozen a year, and the fluidic assembly takes 2 days to bond together, and another day to pass or fail test, so you always have a stock for testing) that is acceptable and not going to delay any deliveries.
$400,000
and two badge wires
Yes that is actually very impressive. I have seen more in handheld fluke DMM:s back then.
I heard "which is designed to extract souls from a sample then treat them with a number of different chemicals"...
This is gonna be good :D
That is some very nicely built kit
What happened to your poor camera, loads of blobs that obscure details in the image. At least you seemed to have cleaned the lens later on.
Better to have biotech going "apple" then apple going biotech ...
That optical block is probably made from Invar, as it has almost zero TC, so the focus does nor change with temperature, and also will nor change with the top being cooled and the base being heated.
Invar is magnetic though, and Mike said this isn't. I had the same thought initially!
There are similar instruments that actually use 300mm-wafer image sensors, a single sensor per wafer. Designed by companies such as Caeleste in Belgium.
"Wankery". Precisely.
The attention to detail at 24:40 is incoherent with the "wankery" at the beginning.. almost as if two separate designers were at work, one for the guts, and another for the lipstick.
Spotted a Qioptiq logo there, makers of very high end commercial and military-spec optical stuff. I guess all the fancy visual LED and metal machining is for when the venture capitalists get a visit to the latest company they spun out. They can see how their money was spent!
this is great, I can''t believe you got that for 120 squid
No wonder it was 400.000 US$, just imagine how many people with specialised knowledge were engaged in developing, designing and building these things and its not like they sold in millions.
Looks like it was designed by former audiophile equipment company dudes 😅
Thanks for the video, it's really interesting to see these teardowns of very specialised technical equipment. Probably not cheap (for you) to buy either.
Ah! Reminds me of the failed attempt from eddify, in a different yet also niche field (NDT) to do something similar. Plus, they even added subscriptions to use the hardward, and have it bricked if you dont renew :)
Teardown time is my favorite time! Nice one Mike👍
I think it appropriate that the external case and monitor should somewhat be aligned with the internals. Would it make sense to have the external body/interface be minimal (read:cheap/standard dull) when it is
Company to their engineers: We need a medical machine but it can not be to expensive but we give you all free hands so you can do whatever you want to!
The Engineers:
Never say that to an engineer, this is what would happen..
I just hope this wasn't bought with NHS funds. Their suppliers already regard them as the ultimate cash cow
The nominal reading of the first sentence is slightly at odds with the instrumental incentives implied by the second.
To be fair, I have worked with several CPOs that thought "It has bluetooth" is a reason to buy something.
Makes you wonder how much of the $400K pays for the marketing wankery?
Give me the purely functional version.
Apple level aesthetic wankery. Classic.
The dirty lens is really aggravating my ocd.
I love engineering undertaken without the control of the accounts department. It looks like they could halve the bill of materials without effecting the functionality. Good score Mike. Thanks.
400k and the touchscreen barely knows what a touch is
Awesome! Its a dream to make project of machine like this one
There exist many entities that thrive on generous government grants by finding ways to spend ever more money on such things. In such cases the high price becomes a welcome feature.
Name 3
@@NeungView You know what... I must have been hallucinating. None of the entities in biotech, healthcare, pharma, academia and certainly not in the defence industry who receive government grants would _ever_ seek novel ways to spend large sums of money in order to justify increases in future funding. My bad!
Ahh, the Polaris.. 🎶 Hey, signal strange. You're lookin' happily deranged... 🎵
There are green PCBs in there. Cant be from apple (:
Definitely resin printing behind those translucent fluid manifolds. Health insurance markups & a quantity of 1 would make it cost 1/6 of a house. Guess we'll have to buy $400k of mandatory health insurance & the net investment income tax won't be tracking inflation.
What normal house costs $2.4 million?
Good score, find these recent devices really interesting. More $100k+ devices that can be dismantled for cheap the better!
You should talk with @LOOKMUMNOCOMPUTER about making a teeeeny tiny pipe organ out of this thing. Or rather an electric accordion.
I can't disagree with most of these comments, especially the "Apple-level aesthetic wankery".
OTOH, If your health-care team (and mine) are relying on machines like this to diagnose what is wrong with you,
or to analyze what kind of therapy is most effective to make you better or keep you alive, I am somewhat reassured
at the level of insane build-quality and intricate custom manifolds, etc. and automation of what was probably a rather
fiddly series of laboratory tests.
This illustrates how much mechanical/electrical/chemical/billogical/optical automation is taking over things that
literally affect our life, health, and safety. And when A.I. starts reading our medical charts and talking to machines
like this, we are handing over the welfare of human society to the machines. Let's try to not interface things like this
to "Skynet" (or SKYNET). 😲
There is a supplier for Electric and Phneumatic PCB near Hamburg. Saw their stuff at the SPS show in Nuremberg in 2016. So it is available, but the number of layers in this manifold is extreme. Although as you say a simple thin multilayer, or even a flexible foil PCB could have done a better job than the extremely expensive wire loom. But looking at the overall concept of this machine, there was little mind given to optimization. At a price of 400k there was surely a fair amout of profit in this. So, inguessbthete is a lot of scope to bring that kind of device to a much lower price point by a less lavish implemebtation an the application of a little judicious production optimization.
I remember seeing well plates at a Medica show in Duesseldorf in the early naughties, but I never saw a machine that drove them.
How did you find sth like that in the first place?
I wonder how they design those microfluidic stacks. Perhaps they just use regular PCB layout software with special footprints, extract the different layers from the gerbers, edit some stuff manually and use those for machining the plates before bonding?
what? lol, why would you use altium, it would make zero sense, you can be sure there is specialized CAD sotware to do that, if not they just used industry standards like Katia or solid edge.
@@lo2740 I never mentioned altium specifically.
And why exactly would you not use PCB layout software, why does it not make sense? The whole stack is just a 2.5D problem, no need for full 3D modelling. Essentially these stacks are also just traces on substrate. Instead of copper it is flow channels (lack of material) and instead of substrate you have poly carbonate. The output of PCB software can be chosen as vector files, and those are perfectly fine inputs for CNC routers. If you can CNC the plates, and stack them together, you have your microfluidic stack. You tell me why that 'makes zero sense'.
I'm sure there's extensions for Solidworks etc, but it doesnt make it inviable or dumb to think of other ways to do that same task, perhaps even more efficient.
You need the threaded holes and other features which aren't used in PCB CAD. So solidworks, katia, etc. as mentioned would be used. You can also use them to do fluid simulations, calculate tolerances, part weight, and more.
Heck of an image sensor. It would be interesting to see what that is capable of. Shame there's probably not a lot that can be done with it otherwise without a whole bunch of very custom work. But you can definitely see where all that ungodly weight came from. All that metalwork for the optics and whatnot was pretty wild to see. Though I can safely say that the external case just looked stupid to me. Probably an absolute crapton of machining to make it look like a grey block cut out from a bigger chunk of something by someone using a one inch drill bit. Strange.
Probably easy enough to interface, USB, though likely a lot higher resolution than what they record. Probably used because they have a higher end version that uses it, so use the larger sensor, and crop the image. Makes alignment easier though.
There is a video out on it now and it was actually very easy to interact with it since there existed a simple program that could control it.
Wow they really gilded the lily on that one, for no good reason. Seems like a mechanical engineer had a wet dream or something on this thing. The expense paid for some of it's features is really comical (like how they did all that just for a fancy door movement).
"Wankery" is such beatiful and - in these cases - useful word!
"We spared no expense!" John Hammond Jurassic Park.
*(that's why this thing costs the same as a 50-seat Provost Bus coach!)
26:29 That aluminium piece on the top left was made from a solid chunk of aluminium.. I mean FFS!
The heavy frame maybe made from cast stainless steel.
Biotech is trending towards appleization. Hgih end (millions usd) equipment with minimalistic interfaces, mechanics so complex you need a specialized technician for anything, and propietariy softwares that are often hard to obtain. Difficult to do science in poor countries
Could the large, heavy metal optics chassis be a non-magnetic grade of stainless steel? Also could be zinc.
Next video coming soon? :)
14:45 - what a takedown! 😆
Unbelievable. I guess some biotech group couldn't swing buying a company Rolls and had to settle for a fancy bit of test equipment they didn't really need.
$400,000 made me gag
Edit: $1,000 each made me gag