Author Topic: Estimating costs for an end-to-end project  (Read 1283 times)

0 Members and 1 Guest are viewing this topic.

Offline electrotwelveTopic starter

  • Contributor
  • Posts: 40
  • Country: in
Estimating costs for an end-to-end project
« on: November 05, 2020, 09:29:21 am »
Hi, How would you estimate costs for an end-to-end low-volume production project (approximately 100 in quantity) ? It includes design, prototyping and then eventually manufacturing the batch too. What are the typical industry numbers for the design phase when done in countries such as US, Canada, UK and Europe vs in countries such as China, India, Taiwan, S. Korea etc?
 

Offline RoadRunner

  • Frequent Contributor
  • **
  • Posts: 397
  • Country: de
Re: Estimating costs for an end-to-end project
« Reply #1 on: November 05, 2020, 10:39:13 am »
Same way you would estimate a how much a house cost.
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 9327
  • Country: fi
Re: Estimating costs for an end-to-end project
« Reply #2 on: November 05, 2020, 06:03:44 pm »
Even for a very rough approximation, we would need to know what kind of product?

Some LED blinker might be some tens of hours to design and get into manufacturing. A precision tool could be 1000s of hours.
 

Offline olkipukki

  • Frequent Contributor
  • **
  • Posts: 790
  • Country: 00
Re: Estimating costs for an end-to-end project
« Reply #3 on: November 05, 2020, 07:00:17 pm »
Same way you would estimate a how much a house cost.

estimated house cost x 2 = actual house cost?  ^-^
 

Offline Brumby

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
Re: Estimating costs for an end-to-end project
« Reply #4 on: November 05, 2020, 11:38:12 pm »
Same way you would estimate a how much a house cost.

actual house anything cost = 2 x estimated house anything cost  ^-^
FTFY

I have found this to be so true in a lot of situations.  It has one condition, though - your estimate must include everything you can think of to produce a number that can be defended from all criticism, so that everyone can agree it is valid and accurate.

In reality, it will just be out by that factor of 2.
 

Offline Brumby

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
Re: Estimating costs for an end-to-end project
« Reply #5 on: November 05, 2020, 11:42:23 pm »
Even for a very rough approximation, we would need to know what kind of product?

That, and a dozen other basic questions, followed by hundreds more.

The question is unanswerable in its current form.
 

Online themadhippy

  • Super Contributor
  • ***
  • Posts: 3264
  • Country: gb
Re: Estimating costs for an end-to-end project
« Reply #6 on: November 05, 2020, 11:47:57 pm »
Quote
In reality, it will just be out by that factor of 2.
unless its a government contract,for those  Actual cost=estimated cost^ (estimated completion  date x actual date of completion) +  a bit more
 

Offline srb1954

  • Super Contributor
  • ***
  • Posts: 1124
  • Country: nz
  • Retired Electronics Design Engineer
Re: Estimating costs for an end-to-end project
« Reply #7 on: November 07, 2020, 03:29:42 am »
Same way you would estimate a how much a house cost.

actual house anything cost = 2 x estimated house anything cost  ^-^
FTFY

I have found this to be so true in a lot of situations.  It has one condition, though - your estimate must include everything you can think of to produce a number that can be defended from all criticism, so that everyone can agree it is valid and accurate.

In reality, it will just be out by that factor of 2.
An estimation error factor of 2 or less is generally only achievable by someone who has had decades of experience in the electronics industry.

If you are just starting in electronics then you should be looking at an error factor of 5 or, if it is primarily a S/W based product, then a factor of 10 is more appropriate when estimating the development time.
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 28429
  • Country: nl
    • NCT Developments
Re: Estimating costs for an end-to-end project
« Reply #8 on: November 07, 2020, 03:49:39 pm »
Actually it is not so difficult for as long as all the requirements are known. The trick is to split a project in many small chunks (including testing!) and do a solid (based on experience or research) time estimation for each chunk. For well defined projects I'm less than 10% off in most cases.
« Last Edit: November 07, 2020, 03:52:36 pm by nctnico »
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 9327
  • Country: fi
Re: Estimating costs for an end-to-end project
« Reply #9 on: November 07, 2020, 05:04:54 pm »
The fixed 2x error factor works for those who are experienced enough to be able to enumerate everything that goes into project and avoid most surprises and problems (and predict the amount of time used to solve "surprising" issues).

Then, obviously, the fact that there is this mystic 2x factor to begin with, is just ridiculous. But at least for me, it is there; it's psychology! Even if we are experienced, we are not always fully sensible.

In theory, if someone is able to estimate the time with such consistent 2x error factor, nothing prevents them from just making a correct estimate to begin with; just multiply by two. In theory. But psychology comes into play. We think we have already multiplied it by two, do we need to do it again?

Clients also want to hear optimistic schedules. Many even understand and accept that project timelines stretch. If you give them the honest timeline, they add 2-3x on the top of it, and say "no" to the project. Yet as a designer, you need to be honest and cannot safely assume such behavior. I have solved this by being more verbose about the timeline so that it's not a single number, but we reach at least some kind of understanding how the project actually works out.

It's all psychology.
« Last Edit: November 07, 2020, 05:06:44 pm by Siwastaja »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf