Cost Estimate

[AerialAgriculture]

Hi everyone!  I would like to get some bids on an integrated circuit design.  Most of the hardware we are using is open source.  The system design includes 6 CCD, MCU,  Direct Memory Access,  GPS unit,  power regulation, clocks, accelerometers, compass, temperature, pressure, and SD card.  We have most everything bread boarded up.  We need an expert to look through the parts and reduce the BOM and produce gerber files for manufacture.   If your not interested but are familiar with pricing I would like to hear your opinion!

Thank You
Modify message

[retrolefty]

Final cost and completion date will be at least double first estimate.

[AndyC_772]

That may well be true if the spec for the product changes during the design process, or it's not properly defined upfront. The key to keeping a job on time and budget, is to have already worked out exactly what's needed before you ask for a quote, and most importantly, don't change your mind afterwards. Ripping up work and doing it again is not free, and any question of the form "can you just...?" is likely to carry a substantial cost penalty.

A good engineer will ask plenty of questions about your project before asking for a quote. There's the obvious stuff like, what does it do, and which parts you've already decided you want to use. Then there's the less obvious stuff, like the environmental conditions it needs to withstand, the volume in which it will be produced, the regulatory requirements that it needs to meet, and the level of support you require (documentation, production testing, integration with other systems). You don't necessarily need to know all the answers straight away, but a good engineer will help you work them out before producing a quote.

Personally, I prefer not to get too involved in jobs where much of the design work has already been done by someone else. If I'm going to end up taking responsibility for whether or not the finished product actually works, I still need to do a thorough design review to spot any bugs. Doing this properly can take nearly the same amount of time as doing the design from scratch.

I clearly recall one job in which the schematics had already been drawn up by the customer but they were buggy, and despite my protests, I was specifically told to design the board 'as-is', bugs and all. So I did, and the board didn't work properly. Yours truly got the blame, of course, so now I tend to only quote on jobs which are of the form "we need a circuit to do X", rather than, "we have a prototype that does X but need it turning into a proper circuit".

[nctnico]

Sounds like a job which is more than just checking the schematics. Putting off the shelve blocks together on a breadboard for a demonstrator is one thing but creating something which is reliable and doesn't rely on parts/sub assemblies which may or may not be available is a different story. I'd be interested but I'd go the same route as Andy: start with getting the requirements on paper and design a circuit which matches those requirements.

[AerialAgriculture]

I agree. and we can discuss the full requirements off line.  We are an experimental physicist, Electrical Engineer, DevOps, Front end developer.  We have a very well defined system, and we will work with you the assure the clearest communication and requirements.  We have all designed circuits, but I am the only one to take a breadboard to a 4 layer PCB.  So I know more or less what we want.  I would be happy to discuss this project!   

[retrolefty]

From your name I'm guessing we are talking a drone to replace crop duster airplanes?

 

[AerialAgriculture]

We are a precision UAV service company.  Offering cutting edge optics, cloud computing, and analysis to service multiple markets in the agribusiness sector.  We have one of the first UAV designers in the US designing a custom solution with our own patent pending optical system and cloud based analytics.  We are in a high growth market with lots of interest. 

[retrolefty]

We are a precision UAV service company.  Offering cutting edge optics, cloud computing, and analysis to service multiple markets in the agribusiness sector.  We have one of the first UAV designers in the US designing a custom solution with our own patent pending optical system and cloud based analytics.  We are in a high growth market with lots of interest.

 Have the latest FCC FAA (duh) 'drone regulations' been a help or a hinder to your plans?

[Zman]

it would be useful to indicate the country

[AerialAgriculture]

We are USA based and the regulations havent hindered a bit. It will only get better by 2017!

[jlmoon]

 Have the latest FCC 'drone regulations' been a help or a hinder to your plans?

I believe the FAA are the regulatory agency for UAVs.

[gani]

Hi everyone!  I would like to get some bids on an integrated circuit design.  Most of the hardware we are using is open source.  The system design includes 6 CCD, MCU,  Direct Memory Access,  GPS unit,  power regulation, clocks, accelerometers, compass, temperature, pressure, and SD card.  We have most everything bread boarded up.  We need an expert to look through the parts and reduce the BOM and produce gerber files for manufacture.   If your not interested but are familiar with pricing I would like to hear your opinion!

Thank You
Modify message

hi,

assumptions:
1- hardware design schematic is ready. ( they are done by standard Tools such as Altium /  Eagle )
2- MCU programming already done and tested and bug fixed.( coded in standard tools as: IAR EW , KEIL , .... )
3- design and mfg and assemble and test of PCB  should be done and is requested.

estimated time for your project ( for PCB design and test ): 
- check and understanding schematics :
CCD : 20h
MCU : 20h
DMA!: 20h
GPS  : 20h
PWR : 10h
CLK  : 00h
ACCL: 10h
CMPS: 10h
TEMP:  5h
PRSR: 10h
uSD :  5h
others : 20h

total: 150h

- gathering PCB Lib Components + Making some PCB LIB parts:
20h...100h

- Designing PCB on 2Layer / 4Layer , check , modify ...
80h...200h

- PCB sample order and tracking...
5h

- Components provision ( you have already done )

- assemble + pre tests + final tests + verify
50h+50h+25h+25h
= 150h

====================================

(min) grand total of times = 405h
(max) grand total of times = 605h

average time = 500h

cost ( done by professional(10~20y experience) for mass production ) = 500h * $20/h = $10000

cost ( done by mid-level engineer(3~5y experience) for pre evaluation ) = 500h * $10/h = $5000

regards

[gani]

Hi everyone!  I would like to get some bids on an integrated circuit design.  Most of the hardware we are using is open source.  The system design includes 6 CCD, MCU,  Direct Memory Access,  GPS unit,  power regulation, clocks, accelerometers, compass, temperature, pressure, and SD card.  We have most everything bread boarded up.  We need an expert to look through the parts and reduce the BOM and produce gerber files for manufacture.   If your not interested but are familiar with pricing I would like to hear your opinion!

Thank You
Modify message

assumption B:

1- hardware design schematic is ready. ( they are done by standard Tools such as Altium /  Eagle ) + revision for reducing BOM :
2- MCU programming already done and tested and bug fixed.( coded in standard tools as: IAR EW , KEIL , .... )
3- Revision + design and mfg and assemble and test of PCB  should be done and is requested.

add 100h~300h to previous post of mine,

average time = 500h+200h = 700h

cost ( done by professional(10~20y experience) for mass production ) = 700h * $20/h = $14000

cost ( done by mid-level engineer(3~5y experience) for pre evaluation ) = 700h * $10/h = $7000

=============================================================

I have been in design and production process of many systems from 1995 till now ( 20 years )
starting by MCS51
.
.
.
Now on ARM LPC17xx , LPC4xxx , AT91SAM , STM32F4XX ,
and RENESAS RX , RL78

and I have designed many measuring and logging tools for  physical/chemical identities ( pressure , vaccuum , Temp , Magnetometry , Load , Flow , Speed , ....)

So , I do exactly know what is your project and how many man-hours it need's to complete and industrialize it,

regards

ganigangi@gmail.com

========================================================================

note:
please keep in mind that, for preparing a design to mass production, many preparations should be done,
so the experienced engineering team can do these process in lower time,

[nctnico]

I'd say you work too slow and too cheap. Sounds more like a (at most) 300 hour job to me. Maybe less if the open source software isn't total crap.

[AndyC_772]

As I interpret it, there's no software component to what's being requested, just a consolidation of the hardware into a single production ready unit.

My main concern on the software side would be how to test the completed hardware, and how to obtain reliable diagnostic information if the only software available to run on it is open-source code that the board's designer hasn't been involved with.

[nctnico]

Perhaps but it would be better to provide at least a reliable software layer which talks to the hardware to prevent a circular discussion about whether it is a bug in the hardware or software.

[gani]

I'd say you work too slow and too cheap. Sounds more like a (at most) 300 hour job to me. Maybe less if the open source software isn't total crap.

good,

Please tell us some examples of your works and projects , if it is OK,

I want to re evaluate my engineering speed and  evaluations

1- project name and functionality   and your previous experience on it
2- subsystems
3- MCU type
4- hardware design process time
5- sch and pcb design and sample assemble time
6- related standards and regulations study time
7- Quality planning time + laboratory test planning and tester provision time
8- software coding + debugging time ( your own codes , ready to use libraries separately )
9- field test planning and tester provision time
10- certificates and approval process time
11- pre prodution line setup
12- production line setup
13- mass production volume

===================================================

the price and cost that I mentioned is an average from US-EU-AUS-NZ-INDIA-CHINA-JP-KOREA

for US and WEST EU , the factor of 1~4  times for cost will be effective
for AUS-NZ-JP-KOREA  , the factor of 1~2  times for cost will be effective
for -INDIA-CHINA , the factor of 0.2~0.5  times for cost will be effective

===================================================

and please tell us your "software coding + debugging" speed in C for popular MCUs ( ARM , TI , NXP , ST , PIC , ... )

( LINES PER DAY )


regards

[AndyC_772]

Gani, only you can possibly know how long it takes you to complete various tasks. It's not something you can look up; you have to rely on your own personal experience and quote accordingly.

My feeling is that there's nowhere near enough information posted here to give a reliable estimate of how long this specific job will take. Without seeing what exists already, I wouldn't want to make any assumptions about how close either the hardware or the software are to being production ready, or what additional work will really be needed to get this product completed. Bear in mind, it's not the customer's job to be expert enough to know the difference between a lash-up that works on the bench, and something that's ready for sale.

The hourly rates you've mentioned wouldn't apply in the US or western Europe. It would be helpful to know which country you're in, so we have context.

I also don't think it makes sense to mention two different prices to have the job done by engineers with different levels of experience. The job needs doing to a given standard, and it's up to the customer to specify what that standard needs to be. Given equal requirements, I'd expect the senior engineer to need fewer hours than a junior one, which of course goes some way to justifying a higher hourly rate.

[gani]

Dear AndyC_772,

thanks for you notes.

please send your evaluations on this specific project (this topic subject).

( time hours + hourly cost )

regards

[Spikee]

I see these kinds of project descriptions often and these cause me to avoid those because of the non-clear specifications.

I would advice the OP to make a document of what is needed and what is already achieved for an engineer to do a rough assessment of what is done and what has to be done.

What often happens is that the client has a very rough idea of what has to be done or does not know what has to be done. This causes long and expensive delays finding out which functionality and hardware the client actually wants...

Also waiting for decision for example which wireless technology or module to use until the project is basically done is often a headache for the designer. While a proper contract gives way for the designer at that kind of point you can't just stop or whatever ...

Make a document , show it to multiple designers / design firms and hear back what you can expect.
I would also advice you not to just pick the cheapest one and this can backfire really easily.  Pick the one that has done something similar or at least have some kind of track record via portfolio or whatever.

I wish guys the best of luck!

[AndyC_772]

please send your evaluations on this specific project (this topic subject).

( time hours + hourly cost )

My feeling is that there's nowhere near enough information posted here to give a reliable estimate of how long this specific job will take.