-
How does a hobbiest go about getting a product designed?
Posted by
BrendonV
on 31 Jan, 2017 08:09
-
Hi,
I was hoping for some advise from those in the know. Basically I have a design and some documentation on an industrial control product I would like to develop for a niche application. Whilst I love to play around with electronics as a hobby designing a robust, safe and quality product is beyond me. I am also limited in the capital investment I can put into the product. I'm stuck in the sense that I have a clear vision and design for the product but cannot afford to just take that to a design firm and say there you go due to the huge investment. Any tips from any EE out there about other ways to bring my design to reality? Is freelance an option? Is there a way I can document my design and send it to company's to tender on?
Excuse me for keeping the product idea to myself at this point as it is relatively niche to a specific industry.
Thank you for any guidance and advice you may have.
-
#1 Reply
Posted by
danadak
on 31 Jan, 2017 14:12
-
If you have a basic proto you can always consider selling design outright
or negotiate a royalty. In order to be successful normally you need some
IP that is scarce, eg. attracts investor because of time to market, proprietary
considerations.
It helps if you have done some market research to establish to an investor that
your ideas are not a one off product. Also your vision must be well expressed,
usually in the form of a business plan, even a minimalist plan has value.
Regards, Dana.
-
#2 Reply
Posted by
AndyC_772
on 01 Feb, 2017 20:18
-
The issue is that if you identify yourself as a "hobbyist", this implies that you see yourself as a member of a group which is typically time-rich, cash-poor, and for which learning new skills and making stuff yourself is the norm. Hobbyists don't get products designed for them; they do the design work themselves.
The alternative is that you are a business, with the resources to get a product developed, tested, approved, and put into production.
The distinction between the two may be entirely a matter of commitment and perspective - but you do need to decide which camp you fall into.
If you don't have the skills and knowledge to develop your idea yourself, then you need to hire people who do have those skills at normal commercial rates. If you can afford to do that out of your own means, then that's great.
If not, you may need to go to your bank with a business plan, and take out a loan to cover the start-up costs for your business, which will include R&D costs, insurance, testing, approvals, tooling, manufacturing, marketing, advertising, and so on. That's what banks are for; to assess whether or not a business proposition is likely to generate a profit, and to fund or decline it depending on the outcome of that assessment.
It may be that you can partner with an engineer who will provide skills, time, effort and IP in return for a stake in the business. The size of this stake will, of course, have to reflect the true commercial value of the work which the engineer has done, multiplied by a factor to account for the risk that the project may not be a success.
-
#3 Reply
Posted by
picandmix
on 01 Feb, 2017 20:41
-
Maybe the " niche application " is just too niche.
If a true industial application it probably would have been developed before, if worthwhile.
Forgetting any R & D, Marketing etc , have you done any basic costings as to the actual profit per unit sold and how many units you could likely sell in a niche market, 1k ,10k 100k + ?
-
#4 Reply
Posted by
Brumby
on 02 Feb, 2017 05:57
-
If a true industial application it probably would have been developed before, if worthwhile.
This would be my first concern.
If you have not done a search on this idea to see if it has - do it now. You may find it has been done before and either wasn't worth it - or has been adopted already in another form. You might also find IP issues.
If you have done a search and found nothing, then consider whether you have used the correct terminology for your searches. Some industries have developed a jargon which has terms that appear to an outsider as quite illogical, until you trace the origins - back to the industrial revolution, the place it was first used or the inventor's daughter.
Not know what you have in mind makes offering advice interesting - but we can understand your caution.
My curiosity is piqued, however.
-
#5 Reply
Posted by
EEVblog
on 02 Feb, 2017 08:51
-
If not, you may need to go to your bank with a business plan, and take out a loan to cover the start-up costs for your business, which will include R&D costs, insurance, testing, approvals, tooling, manufacturing, marketing, advertising, and so on.
That's the absolute last resort. That's the traditional business school model and it comes with a high failure rate and often massive blow-out's on budget.
If you can't fund it most of it yourself then don't do it.
Put what money you have into a polished prototype and try to attract an order from a company that has a use for it.
Or take that polished prototype and quote it out to "production-ise it".
That's what banks are for; to assess whether or not a business proposition is likely to generate a profit, and to fund or decline it depending on the outcome of that assessment.
Banks know jack-all about technology, and even less about niche industrial products.
Just because look over your proposal and give you money means almost nothing about it's actual real viability.
You could write a polished BS business proposal for a perpetual motion machine and they would fund it if the presentation is good enough.
Heck, companies like Juicera get 10's of millions of dollars in funding from
professional tech investors for the dumbest product idea in history.
http://liesandstartuppr.blogspot.com.au/2016/08/how-is-this-thing.htmlIt may be that you can partner with an engineer who will provide skills, time, effort and IP in return for a stake in the business. The size of this stake will, of course, have to reflect the true commercial value of the work which the engineer has done, multiplied by a factor to account for the risk that the project may not be a success.
This is a reasonable option, especially if you have a decent polished prototype and all they have to do is productionise it.
Don't go to an engineer or freelancer with just and idea, they will either say "thanks but no thanks" or charge you an outrageous sum because of the risk involved.
Good freelance engineers and consulting houses are experts as risk assessment on whether a project will be a big PITA, that's their livelihood.
-
#6 Reply
Posted by
EEVblog
on 02 Feb, 2017 09:22
-
If a true industial application it probably would have been developed before, if worthwhile.
This would be my first concern.
If you have not done a search on this idea to see if it has - do it now. You may find it has been done before and either wasn't worth it - or has been adopted already in another form.
I've seen this countless when people contact me to help out with their brilliant idea.
In most cases the idea is not original or obviously will not be viable because *insert reason*. Often the simplest of google searches with the right terms will reveal this, yet people seem incapable of doing this. Sometimes it's an industry knowledge thing.
-
#7 Reply
Posted by
EEVblog
on 02 Feb, 2017 09:26
-
If you have a basic proto you can always consider selling design outright
or negotiate a royalty. In order to be successful normally you need some
IP that is scarce, eg. attracts investor because of time to market, proprietary
considerations.
In the industrial space you simply need a product that solves a problem for the client.
Very common to show a prototype and they will order 10 or a 100 of them. Use the profit on those to bootstrap your business. Let your client/customer be your investor.
It helps if you have done some market research to establish to an investor that
your ideas are not a one off product. Also your vision must be well expressed,
usually in the form of a business plan, even a minimalist plan has value.
Only a fool invests in an idea and business plan. A nice working proto is worth 1000 business plans.
-
#8 Reply
Posted by
AndyC_772
on 02 Feb, 2017 09:39
-
If not, you may need to go to your bank with a business plan, and take out a loan to cover the start-up costs for your business, which will include R&D costs, insurance, testing, approvals, tooling, manufacturing, marketing, advertising, and so on.
That's the absolute last resort. That's the traditional business school model and it comes with a high failure rate and often massive blow-out's on budget.
If you can't fund it most of it yourself then don't do it.
I guess you could substitute "bank" with VC, friends, family, or anywhere else you can think of who might be prepared to help fund the project. The point I was trying to make, is that developing a product that's fit for sale does require an upfront investment, which has to come from somewhere.
It may be that you can partner with an engineer who will provide skills, time, effort and IP in return for a stake in the business. The size of this stake will, of course, have to reflect the true commercial value of the work which the engineer has done, multiplied by a factor to account for the risk that the project may not be a success.
This is a reasonable option, especially if you have a decent polished prototype and all they have to do is productionise it.
I'm sure you'd agree that "all they have to do" could mean anything from laying out a PCB the right shape, to a complete redesign.
A working proof of concept may indeed be a good way to secure backing for a project. It may, however, be unwise to assume that much of it can be turned straight into production-ready hardware.
If nothing else, the engineer producing the final version will be responsible for it, and that can involve a lot of design verification work. "Why did they do X?" can sometimes take as long to answer as "how would I achieve X"?.
-
#9 Reply
Posted by
Psi
on 02 Feb, 2017 10:21
-
Design it all yourself.
You can get a professional looking product with a nice off the shelf extruded aluminium enclosure using some PCBs as front/rear panels. This gives you the correct shape panels, with holes in the exact place you need to mount switches/knobs etc
You also get silkscreen for text/labels and you can even use gold plating to make gold text/labels as well.
Get an extruded enclosure with slots down the side for holding the PCB.
Hammond have some and will do custom orders for any length.
You can use 90deg headers for mounting a little PCB vertically behind the front PCB if you need to mount tact buttons or leds on. Or if your cheap use the slot method where the pcbs fit into routed slots and you solder the tracks together.
-
#10 Reply
Posted by
EEVblog
on 02 Feb, 2017 11:40
-
That's the absolute last resort. That's the traditional business school model and it comes with a high failure rate and often massive blow-out's on budget.
If you can't fund it most of it yourself then don't do it.
I guess you could substitute "bank" with VC, friends, family, or anywhere else you can think of who might be prepared to help fund the project. The point I was trying to make, is that developing a product that's fit for sale does require an upfront investment, which has to come from somewhere.
But that upfront investment doesn't have to be as significant as you make it out to be. Nor do you need to do all the things you mention "R&D costs, insurance, testing, approvals, tooling, manufacturing, marketing, advertising, "
All you may need is some R&D and initial prototype costs.
Once you have that prototype to demo then you can shop it around and get essentially pre-sales.
Marketing and advertising are a huge guaranteed waste of money if you are targeting a niche industry app like the OP is.
You could argue insurance too and all the other usual stuff that goes along with operating a professional business.
If you go into the game thinking that the only way is to follow the MBA business playbook, then you are doing yourself a disservice.
It may be that you can partner with an engineer who will provide skills, time, effort and IP in return for a stake in the business. The size of this stake will, of course, have to reflect the true commercial value of the work which the engineer has done, multiplied by a factor to account for the risk that the project may not be a success.
This is a reasonable option, especially if you have a decent polished prototype and all they have to do is productionise it.
I'm sure you'd agree that "all they have to do" could mean anything from laying out a PCB the right shape, to a complete redesign.
A working proof of concept may indeed be a good way to secure backing for a project. It may, however, be unwise to assume that much of it can be turned straight into production-ready hardware.
[/quote]
Sure. But when you have a functional prototype the unknowns (and hence risk) are almost always much fewer. Lower risk and less unknowns means a more accurate quote and a choice of more people who are willing to take it on.
-
#11 Reply
Posted by
AndyC_772
on 02 Feb, 2017 11:51
-
If there are things you know you don't need, then that's great
Much better that, than to discover further down the line that there are things you hadn't thought of
-
#12 Reply
Posted by
BrendonV
on 02 Feb, 2017 13:17
-
Thank you all for your words of wisdom it has helped me think through the process. Yes i have done some searching and this product does exist in may forms however not in the form of which I plan to implement. It is hard to comment any further without going into detail. I have been in this particular industry for 20 years and innovation in this particular control product hasn't changed since the 70's. The existing hardware in the industry is hard to use, clunky and lacks the many features we can utilize in modern day electronics.
I feel my best path is to first develop a prototype to physically show the functionality of the product. My question would be... how would an engineer / consulting house react when i walk in with my "beginner hobbiest" prototype and ask them to turn it into a commercial product. Do i look like a time waster? How good does my prototype have to be to be taken seriously? Sure i can put a development board in a box, connect in sensors buttons and write a program to get a working demo going but is that enough?
Thanks again
-
#13 Reply
Posted by
AndyC_772
on 02 Feb, 2017 14:21
-
You won't be turned away because of the quality (or not!) of your prototype. Engineers expect to provide the design expertise, it's part of the job. A good one will walk you through the process of having bespoke electronics designed for the first time, and once they understand what it is you're looking to achieve, they'll give you a rough idea of the amount of time and effort which is likely to be involved.
They'll also ask a lot of questions about how the product will be used, where, by whom, and what specs it needs to meet in terms of temperature, power, EMC, safety, reliability, serviceability and so on. It's really important to be able to nail these down *before* any design work starts, as changes later on mean throwing work away, and that's costly. A good engineer will help you work out the full specification if you haven't already done it yourself.
For a complex product, there may be a feasibility study to do before they can give you an accurate estimate of time and cost. If it's relatively simple, that may not be necessary.
A proof-of-concept mock-up of the finished product is more useful to other parties, such as potential customers or end-users. An engineer won't need to actually see, say, a dev board hooked up to some I/O devices just to know that a device can be made to work that way.
Where there's a choice, they may also want to use their own preferred parts rather than the ones you've prototyped with.
You may get turned away if an engineer is simply too busy, which often happens with good ones. Some specialise in a particular area, and may be unwilling to take on your project if it requires them to do a lot of background learning which won't then be relevant to jobs they're likely to work on in future.
You'll only look like a time waster if you turn up with a pitch along the lines of "here's my great idea, please will you do all the work, and I'll let you keep a piece of the profits".
-
#14 Reply
Posted by
Brumby
on 03 Feb, 2017 00:47
-
Having a working prototype - however frail - is a big step in the right direction. It should be presented neatly enough so that things are clearly laid out and if it's three times bigger than it needs to be when in production, don't worry. You can show what it needs to do - and, from your own efforts, you will have come across limitations that need to be overcome. Certainly bring your notes and diagrams as these, too will be valuable.
However, I see your biggest challenge will be finding the right engineer.
-
#15 Reply
Posted by
EEVblog
on 03 Feb, 2017 01:31
-
I feel my best path is to first develop a prototype to physically show the functionality of the product. My question would be... how would an engineer / consulting house react when i walk in with my "beginner hobbiest" prototype and ask them to turn it into a commercial product. Do i look like a time waster? How good does my prototype have to be to be taken seriously? Sure i can put a development board in a box, connect in sensors buttons and write a program to get a working demo going but is that enough?
Yes, that will be enough, as long as it demonstrates most the basic functionality you have in find.
Experienced engineers are good as sniffing out how much work is required to polish it up and risks involved in commercialising it.
Also, beware that there are easy ways and hard (read expensive) ways to commercialise a product.
For example, if you want to make only a few thousand of them then an off-the self case is entirely appropriate, and as someone mentioned before, a PCB front panel etc is more than enough to look like a commercial polished product. Nice low risk, low tooling outlay option. But someone might try and convince you the only way is fully custom moulding with $59k in tooling and hiring packaging graphics experts etc etc. Or vice-versa, not advising you correctly if your requirements are high volume.
-
#16 Reply
Posted by
EEVblog
on 03 Feb, 2017 04:34
-
-
#17 Reply
Posted by
wilmer
on 03 Feb, 2017 04:53
-
Hi,
I am PhD student in control
if you want to check your idea, I am willing to hear your idea and if is industrial control I can point you in the correct direction.
Prepare a confidentiality contract and we can discuss your idea without you worrying about it.
cheers
-
-
Almost all products these days are mostly software in terms of development cost. If you can do that yourself, that's a huge start. There are enough resources out there to teach yourself.
Processor power is cheap, so if you need fancy GUIs or complex processing ( e.g. image recognition) etc. it's entirely feasible to use higher-level programming languages and libraries and run it on a RasPi or similar.
If you can get something working on a development board, maybe with off-the-shelf add-ons (Adafruit,Sparkfun etc.), to the point where it is pretty much fully functional, that gives you a platform on which to develop the software.
Buy a nice looking off-the-shelf case, so you have a physical thing you can demonstrate to a potential customer.
This will so far have cost very little other than time - probably well under $1000.
Once you have established that it's worthwhile, the process of turning that into a more refined product is basically just condensing the pile of devboards into a purpose-made design, with minimal software work, maybe adding some ruggedising ( input protection etc.), which is a much more finite and cost-able task for a design house, or a one-man-band consultant. In either case, you really need to get recommendations as there are cowboys and idiots out there, especially in the latter category. If anyone says they can start immediately, walk away - good people are always busy.
Another question is how cost-sensitive is it? Many industrial devices can be shown to have a demonstrable pay-back time becaue of time or efficiency savings.
Maybe an initial small batch made from off-the-shelf boards and modules might be feasible, to get some initial sales to either finance future development, or maybe get the interest of a manufacturer who you could license it to.
-
#19 Reply
Posted by
EEVblog
on 03 Feb, 2017 08:52
-
Almost all products these days are mostly software in terms of development cost. If you can do that yourself, that's a huge start. There are enough resources out there to teach yourself.
Processor power is cheap, so if you need fancy GUIs or complex processing ( e.g. image recognition) etc. it's entirely feasible to use higher-level programming languages and libraries and run it on a RasPi or similar.
This is where modules like these come in real handy.
http://www.4dsystems.com.au/product/gen4_uLCD_70D/A complete polished solution that you can easily program to do automated control.
Impresses the heck out of clients and investors with very little effort, just whack it in a case.
Not great for mass production, but great for industrial type stuff in the hundreds or something, just pass the cost on.
-
#20 Reply
Posted by
Psi
on 03 Feb, 2017 09:16
-
Also, if your really worried about telling anyone about your idea, even after getting them to sign NDA or whatever, you can figure out an alternative story about the industry the product will be used in.
Same features, just being used for a totally different purpose.
This goes a little way to safeguard your design.
If after hearing about your idea and industry application they go do some googling and see "wow, that product is going to make millions" they are more likely to try to get in on the action than if they google and cant see that much of a use for your product in the "fake industry" you give them.
-
#21 Reply
Posted by
AndyC_772
on 03 Feb, 2017 09:57
-
If anyone says they can start immediately, walk away - good people are always busy.
Another question is how cost-sensitive is it?
This seems like as good a place as any to mention some of the questions that anyone thinking of designing a commercial product should have a think about. Here are questions I might ask about a project that was completely new to me, in the approximate order I'd ask them.
- What does it do?
- What, specifically, is the end application? In what location and industry?
- Is it 'high risk' in any sense? (ie. safety critical, or otherwise demanding of a very high standard of reliability
- When do you need it?
- What are all the inputs and outputs? (Buttons, lights, data ports, transducers...)
- How is it powered? What voltage range? Is the supply clean, or subject to noise, transients, drop-outs...?
- Who will use it? What level of experience, training or expertise might they have? Might it be wilfully abused?
- What mechanical constraints apply? (Size, weight, materials, mounting points, cabling, connectors...)
- What (rough) quantity will it be manufactured in?
- Target build cost per unit?
- What environment will it be used or installed in? (Temperature, shock, vibration, dust, liquids, humidity etc)
- What regulatory approvals do you need the finished product to meet? Do you need this process taking care of too?
- How long does the product need to last?
- Is it important that it be straightforward to service and repair? By whom, and with what level of skill?
- Which parts of the system do you need an engineer to design for you, and which are already taken care of?
- Are there any existing components (hardware or software) which you'd like to retain?
- Do you require work to be carried out to any specific standards (eg. use of CAD tools or libraries, or your company's internal processes)?
- What do you want to receive in terms of technical data? (Original schematic and PCB CAD files, source code, Gerber data, binary files)
- What do you need in terms of supporting documentation? (User guide, service manual, full technical write-up)
- Do you require full, outright ownership of the finished IP?
- Do you need prototypes manufacturing? Do they need to be the same as the finished product? Or does it make sense to produce a 'developers version' which is different, but easier to work with in some ways (size, memory, I/O capability, links, headers, test points...)?
-
-
Whilst I love to play around with electronics as a hobby designing a robust, safe and quality product is beyond me.
Ask yourself exactly what skills you think you are lacking, or aspects of the prrocess you don't understand, and see if you can fill at least some of them in, or identify some specific areas you need advice on.
Once you've done that, it may be that you can just hire a consultant/expert for a few hours to answer some specific issues that can then guide you into doing more of the work yourself, to get you a lot further down the road.
-
-
- Is it 'high risk' in any sense? (ie. safety critical, or otherwise demanding of a very high standard of reliability
And for an industrial product, remember that "risk" also means expensive downtime if it doesn't work properly.
Who will use it? What level of experience, training or expertise might they have? Might it be wilfully abused?
Do you need to provide a UI or documentation in different languages?
- What (rough) quantity will it be manufactured in?
This is one of the most important questions, and one that is almost always not examined by inexperienced designers/specifiers.
It has huge influence on the direction the design takes.
For a new product, you should probably start off assuming the minimum viable batch size, and if that sells you can always look at cost-cutting later. If you can't make a profit out of a smallish batch, i.e. you need to rely on cost-optimising, chances are the product isn't viable at all long-term.
- Is it important that it be straightforward to service and repair? By whom, and with what level of skill?
Consider fault diagnosis from day one. Make it easy for faults to be narrowed down - status LEDs, meaningful error messages, logging,test modes etc. In an industrial setting, faultfinding time often relates directly to costs of lost production. Remember that faults may initially be investicated by someone whose normal tools are a hammer and a big spanner.
For example look at any industrial sensor - it will always at least have indicator LED(s) for status and power.
-
#24 Reply
Posted by
BrendonV
on 03 Feb, 2017 12:07
-
Wow thanks guys. Overwhelming response. Got a bit of shock when i pulled up YouTube this afternoon and saw Dave answer my question in a eevblab 😀 So much thoughtful and constructive advice is exactly what I needed. I will keep you posted with how I go on this endevour of mine.
Can't say thanks enough for all of your help!