Industrial costs vs. Do it yourself costs

[Topher925]

I'm working on my idea for my masters thesis which involves using some optical instrumentation to measure temperature. To go out and buy the industrially available instrumentation to do this project would cost more than $4,000. Yet, I have found some conference presentations and progress reports from a national laboratory who built their own equipment for less than $100. When researching this project, I honestly can not see how this instrumentation can cost more than $200. I've seen price margins like this with a lot of laboratory related products. How can manufactures justify such high costs for equipment? I know engineers and technicians are expensive, is it usually the labor that's involved which gives companies the right to charge such high prices on low volume products?

 

[mgb_phys]

The parts cost $100.

Assembly costs $50/hour if it isn't super skilled (includes cost of runnign the workshop etc)
How many years of how many R+D engineers (at $150K/pa) did it take to design it?
This cost is divided up by how many you are going to sell - so it's cheap for an iPod expensive for a scope. How many years of R+D were spent on products that didn't sell - you also have to cover these costs.

Then each unit takes many hours of testing and calibration. The test equipement, calibration standards and the skilled workers doing this are expensive.

Then you have cost of sales and support, how many catalogues do you have to print, how many ads do you pay for and how many trade shows do you have to go to to sell each unit?

Then the dealer will want a 40% markup.

So typically all your cost of building the unit and doing business has to be less than 1/2 the sticker price.

 

[gmax137]

don't forget the cost of liability insurance, or the cost of printing up those sheets that say "do not use the oscilloscope while in the bathtub..."

 

[stewartcs]

And then there is supply and demand...not to mention good ole greed!

CS

 

[Topher925]

So I guess I shouldn't be discouraged to make a somewhat complex piece of equipment myself when I see them being sold for prices 50 times that of what the parts to make it actually costs.

 

[mgb_phys]

If you are trying to build something for yourself/your lab go for it - but I wouldn't assume you can build a business on it.

 

[chayced]

A prime example of markup is fisher stop watches. A $5 stopwatch suddenly turns into a $50+ scientific instrument because of QC and traceability. The cost is not in the manufacturing, but instead in the number of people it takes to guarantee it will work as intended and to cover the liability if it doesn't.

Also remember that producing something in mass is much cheaper than producing small numbers. A tool that costs $100 when produced in millions can cost $1000 dollars when produced in the hundreds, and can cost tens of thousands when only one is produced. The actual components/raw materials used to produce something are often negligible compared to other expenses.

Just something to think about.

 

[Topher925]

If you are trying to build something for yourself/your lab go for it - but I wouldn't assume you can build a business on it.

I'm definitely not looking to be an entrepreneur, just do some research without destroying my budget. The incredibly high prices of some of this equipment had me worried but I feel better now. I think another good example to go along with that watch is computers. I built my PC for $800, 3 years ago, and it outperformed most PCs at my last place of employment which were recently purchased for $3000+.

 

[mgb_phys]

You can probably build it yourself easily, it might take a bit of work to calibrate but it only has to work over the range of inputs you are interested in and it onyl has to work for a short time.
Low temperature stuff is a prime example of this, it's incredibly expensive from say oxford instruments compared to something you can knock up for a one off temperature measurement - you also learn somethign about electronics and sensors at the same time.

ps. If it involves building the electronics get a copy of horowitz and hill "the art of electronics"

 

[berkeman]

I'm working on my idea for my masters thesis which involves using some optical instrumentation to measure temperature. To go out and buy the industrially available instrumentation to do this project would cost more than $4,000. Yet, I have found some conference presentations and progress reports from a national laboratory who built their own equipment for less than $100.

Can you post a link to this $4000 instrument? It would help us to figure out if there might be some "gotchas" in making it yourself. Some optical alignment, coatings, etc. can be difficult to do on a benchtop.

Also, if you make it yourself, will it just be out of discrete optical components on an optical table, or will you be building an instrument into an enclosure?

Finally, will you need to use AC Mains power, or can you use DC Lab Power Supplies? If AC Mains power, do you have experience in building projects that conform to UL safety standards? They are not hard to meet, you just need to know what-all to be sure to include in your design.

 

[Topher925]

Here is an example of one: http://www.neoptix.com/reflex.asp

There are a lot out there but they all range from $750-$15,000 for the signal conditioner alone. The probes usually cost $200-$1000. The instrument I will be building probably won't have an enclosure. Just the probe and circuit board. For projects like this I usually use a PC power supply or a wallwart.

Some work differently, but the one that I will be using will be based of the principles of thermographic phosphors.

 

[chayced]

Forgive me, but I originally thought this was for some sort of IR instrument. After a bit of research on thermographic phosphors its pretty obvious that this is not a DIY for the feint of heart.

You need a pulsed excitation laser to excite the phosphorus then another instrument to measure the light emitted by the phosphorus at intervals where the laser isn't pulsing. You could easily spend more than $4000 on manpower and parts constructing your own. And then you would have to calibrate it for your application.

Is there any reason why you can't use a thermocouple or IR thermometer?

 

[Topher925]

You need a pulsed excitation laser to excite the phosphorus then another instrument to measure the light emitted by the phosphorus at intervals where the laser isn't pulsing

If it was 1995, I might believe you. However, the fiber optic communications industry has grown by leaps and bounds in the past 2 years alone. This can, and is, all done with LEDs and photodiodes now. For example, the work by ORNL that originally inspired this project (notice $10 signal processor): http://www.hydrogen.energy.gov/pdfs/progress04/ivh4_mcintyre.pdf

Many phosphors such as ruby (Cr:Al₂O₃) have decay times in the area of 3ms. Using rough numbers, let's say a phosphor gives of 200mcd at a wavelength of 690 nm, could this not be measured using a PMMA fiber optic wave guide and a Si photodiode? A simple preamp on the photodiode with a quick slew rate being read by a 1MSPS A/D converter should be able to take accurate enough measurements, should it not?

Any opinions on the feasibility of this project are more than welcome. I would certainly appreciate some expert opinions before committing myself to a project like this. The reason I don't want to use thermocouples is because they are subject to EMI, require electrical insulation for my application, and I can not find any that have the ability to give accuracies less than 0.5C with .25 second response times while still being less than 500 microns in diameter. IR measurement is out of the question because I am measuring temperatures of a solid surface in contact with gas and liquid flows.

 

[chayced]

Sorry could only find limited documentation in the couple minutes I looked. This project goes a little over my head so I will pass to those with more expertise. Good Luck!

 

[FredGarvin]

Most of the time, there is not enough time for people to embark on a science project. Even if they do, it will probably cost more in the long run in labor hours.

Let's not forget the learning curve that everyone has to go through when learning a new field. Avoiding mistakes can sometimes be priceless. There are plenty of things I use that would be easy to make my own, but I would never risk a multi million dollar prototype engine on it.

 

[amuron]

Unless you have done something fairly closely related in the past, the learning curve is often filled with gee by the way, its not in the data sheet type things and other gotchas.

Having done the 50:1 margin products myself over the years, there are a lot of things that go into pricing well beyond the tech aspects and actual costs. In some cases, the prices are high, as that's what the market has determined they will be, and the market volume is low enough, there is not enough competition to drive prices downward. Those are nice product niches to be in, as often there are subtle tech reasons as well as business reasons as to why no competition exists. Granted, those subtleties can be easy, or can be incredibly complex to solve. It may pay to allocated a fixed period of time to see how much progress can be made. There is nothing wrong with going down a path for a short period only to cry uncle and go to a purchased version as it turned out not to be feasible. Just set the time limits upfront, to avoid the project becoming a huge detour from what you are trying to do.

As far as the actual project, having done similar things in the late 90's... you likely are correct as far as tech changes and costs involved once the initial r-d was out of the way... the big issue is how much time its going to take to get to that point. The pdf won't load right now, perhaps the site is undergoing maintenance otherwise I'd comment further.

 

[misgfool]

Most of the time, there is not enough time for people to embark on a science project. Even if they do, it will probably cost more in the long run in labor hours.

Yes and fast prototyping is also word of the day. Buying semi-cheap industrial grade equipment allows construction of first models under tough deadlines. Especially if one has to explore different options.

 

[Topher925]

Well, I started ordering parts today so I'm starting down the path of no return. I've been reading a lot of thesis's (sp?) that are similar to my thesis project and I really believe this doable. I just wanted to add, that my original intention was to modify an industrial working product and adapt it to my own needs for this project. But I think in the end building my own instrumentation will save me time and money in the long run while being a lot more educational. I might even add on to this project depending on how things go.

 

[stewartcs]

Well, I started ordering parts today so I'm starting down the path of no return. I've been reading a lot of thesis's (sp?) that are similar to my thesis project and I really believe this doable. I just wanted to add, that my original intention was to modify an industrial working product and adapt it to my own needs for this project. But I think in the end building my own instrumentation will save me time and money in the long run while being a lot more educational. I might even add on to this project depending on how things go.

Good luck!

BTW, that's theses.

CS

 

[FredGarvin]

Well, I started ordering parts today so I'm starting down the path of no return. I've been reading a lot of thesis's (sp?) that are similar to my thesis project and I really believe this doable. I just wanted to add, that my original intention was to modify an industrial working product and adapt it to my own needs for this project. But I think in the end building my own instrumentation will save me time and money in the long run while being a lot more educational. I might even add on to this project depending on how things go.

I envy you in the fact that you do have that option. It will be a great learning experience. Hopefully you don't have to "pay" too much for that education.

 

[Topher925]

Good luck!

BTW, that's theses.

CS

Ah, thanks.

Now that I have started this project, I'm already hitting some road bumps. I am having a lot of trouble getting parts and even communicating with suppliers. I call them up, tell them exactly what I need and ask for a quote. They then tell me to send them an email of what I want. So I do so, then maybe a week later and I get a reply saying they are out of stock of something, they don't do what their website says, or say that there is a $1,000 minimum order. This is beyond frustrating.

Is this something that is common for grad students or universities when dealing with suppliers? When I worked in industry I learned that a supplier WANTS to sell you their product so they can make money. If I was to purchase something or wanted information from a supplier I got it ASAP. Now that I am a grad student its like they don't even care. What is wrong with these people?

 

[mgb_phys]

Is this something that is common for grad students or universities when dealing with suppliers?

It's probably the most useful thing you will learn.
Along with deciding if it's easier to put an order on your own credit card and spend 6motnhs trying to get it refunded by the dept vs. spending 6months trying to get the dept to create an account with a supplier so you can order things.

When I worked in industry I learned that a supplier WANTS to sell you their product so they can make money.

They want to sell lots of things regularly to existing clients. They don't want to spend months explaining something to 'some kid' who wants to use a single item in some non-standard way.
If you can find an individual engineer at a compnay that is happy to work with you and wants to help - cherish them, make sure you get put through to them next time, do your homework so you aren't asking stupid questions and say thank you!

You also get used to begging - especially if you work for a prestigous instiutution or your project has some gee-whizz (space) or noble (curing cancer) connection, you can get around their 10,000 part minimum order and get the dozen you need as a free sample.

 

[FredGarvin]

Is this something that is common for grad students or universities when dealing with suppliers? When I worked in industry I learned that a supplier WANTS to sell you their product so they can make money. If I was to purchase something or wanted information from a supplier I got it ASAP. Now that I am a grad student its like they don't even care. What is wrong with these people?

No. It's pretty much across the board thanks to "lean manufacturing." No one wants to have any stock on their shelves any more. There are not very many technical suppliers that can ship you something in a day or so. And because nothing is on the shelves, they now want insane minimums to warrant putting something into production. Personally I think it is a total load of crap. I don't know how many businesses out there survive when they need to get a supplier working 6 months in advance. I never get those kinds of lead times to do my own designs.

 

[stewartcs]

At the end of the year the manufacturer has to pay taxes on what they have on the shelf so they don't like stocking any more than they have to. However, if you buy it first, then they'll be happy to store it for you in a customer holding area (presuming you buy a lot from them).

CS

 

[Topher925]

It's probably the most useful thing you will learn.
Along with deciding if it's easier to put an order on your own credit card and spend 6motnhs trying to get it refunded by the dept vs. spending 6months trying to get the dept to create an account with a supplier so you can order things.

I'm learning this lesson right now. If I used my credit card everything gets ordered quickly so I'm tempted to put everything on it, but I know its going to take forever for me to get reimbursed. I only have until the end of the year to complete this project and don't feel like sitting around for 6 months and worrying about deadlines.

I guess this frustration is felt by everyone that has projects like this? Its going to take me 6 weeks just to get my photodiodes. The only suppliers I am looking at now are ones that specifically deal with universities and research related institutions. Its nice to have salesman that actually return emails and that can give you a quote in less than 14 days. Anyway, I think I got just about all my parts found now, just got to wait for them to arrive.

 

[misgfool]

I guess this frustration is felt by everyone that has projects like this? Its going to take me 6 weeks just to get my photodiodes. The only suppliers I am looking at now are ones that specifically deal with universities and research related institutions. Its nice to have salesman that actually return emails and that can give you a quote in less than 14 days. Anyway, I think I got just about all my parts found now, just got to wait for them to arrive.

Have you done a risk management plan?

 

[mheslep]

Ah, thanks.

Now that I have started this project, I'm already hitting some road bumps. I am having a lot of trouble getting parts and even communicating with suppliers. I call them up, tell them exactly what I need and ask for a quote. They then tell me to send them an email of what I want. So I do so, then maybe a week later and I get a reply saying they are out of stock of something, they don't do what their website says, or say that there is a $1,000 minimum order. This is beyond frustrating.

Samples! Ask for samples Topher, all free. That's basically how the world operates below minimum quantity orders. No purchase orders, or credit card checks either to slow you down. Worst case is you get spammed by their sales people in the future. If the sales guy won't do it, the application engineer often can. The company pay off for a freebie like this is that they're getting their product into the hands of somebody who may very well be ordering much more than minimums in the future.

Edit: caveat emptor, do the sample requests early. The downside risk is that the sample shipment gets dropped through the cracks, the sales guy leaves, whatever, and you have no recourse - there's no order number or enforcement mechanism for your freebie.

 

[mheslep]

At the end of the year the manufacturer has to pay taxes on what they have on the shelf so they don't like stocking any more than they have to. However, if you buy it first, then they'll be happy to store it for you in a customer holding area (presuming you buy a lot from them).

CS

Yes taxes on capital equipment are part of it, but I think its the kink that idle stock puts on ROI that hurts the most - that is the cost of the money required to stock shelves.

 

[Topher925]

Have you done a risk management plan?

Nope, not at all. I'm an ME not an MBA major.

I've tried asking for samples from all these places and got nothing. Not even samples for a few LEDs. Most companies these days won't give out samples to universities but only to major OEMs and manufacturers. For Example; http://www.marktechopto.com/Sample-Cart/leds-sample-cart-start.cfm?part_number=LC503FPG1-15Q-A3

 

[misgfool]

Nope, not at all. I'm an ME not an MBA major.

These days even for engineers it is typical to consider possible risk factors in projects as a part of planning and figure out ways to mitigate them. Besides being preemptive, when it has been written down, responding to problems is faster. It has little to do with the MBA creatures.

 

[Topher925]

UPDATE:

IT WORKS! I got the basics of a single sensor working and just finished testing it and it works BEAUTIFULLY. So much better than I even anticipated actually. I think I might even be able to achieve much greater accuracy and response time than I had even anticipated. I just got to finish fine tuning the amplification circuit and do some more work on the algorithms but after that I am ready for calibration! w00t!

BTW, my total costs are still about 3% of that of industrial versions of this sensor!

 

[FredGarvin]

BTW, my total costs are still about 3% of that of industrial versions of this sensor!

Even factoring in a cost for your time?

 

[Topher925]

Even factoring in a cost for your time?

I'm a grad student Fred. So maybe like 4%.

 

[mheslep]

Nope, not at all. I'm an ME not an MBA major.

I've tried asking for samples from all these places and got nothing. Not even samples for a few LEDs. Most companies these days won't give out samples to universities but only to major OEMs and manufacturers. For Example; http://www.marktechopto.com/Sample-Cart/leds-sample-cart-start.cfm?part_number=LC503FPG1-15Q-A3

Unfortunate. I thought you were a part time grad student - working for somebody doing a high altitude wind turbine project at some point?

 

[Topher925]

I used to be but not anymore. It was a venture capital project which was having issues with funding. Given the state of the economy and the way the project was headed I decided to be a full time grad student.