Devicer wrote:
> Hi friends,
>
> My office plans to invest in setting up a new embedded system
> development facility.
> Our prime business is industrial desing & manufacturing .The new
> facility will develop industrial automation applications that will be
> piggybacked on our company's products.
> So we are starting from scratch.
>
> The applications will be industrial control like material handling or
> closed loop valve control etc., using typical sequnce logic and/or PID
> loops.(What i mean is it is not safety critical/ mission critical like
> missile control automotive safety control. )
>
> I am given the task for preparing a proposal & roadmap for this
> project.
>
> this task includes
>
> identifying resource requirements
> identifying skill set requirements
> gap analysis
> preparing specifications for development environment
> identifying business partners
> preparing procurement notes
> fixing up contracts
> training HR
> project reporting
> handing over the facility to development team.
>
> The time horizon is 6 months from start date
>
> I am looking for some help from the forum in successfully completing
> this project.
>
> I started out like this
>
> Resource reqmts :
>
> 1) Infrastructure
>
> built area, no of seats (approx 5),amenities (power,water,recreation)
>
> 2) Development setup
>
> Architecture (undecided yet) - 8 bit,16 bit,32 bit
>
> Indep Software Devpt tools : Source code editor,compiler
> (cross-compiler?)
> Dependent s/w devpt tools : assembler linker simulator
> Hardware tools :incircuit emulator,device programmer
> Test & meaurement:Scope,Logic analyser,Function
> generator,Timer/counter,Datalogger
> Power supply:AC/DC
> RTOS
?)
> Training boards :
>
> 3) Skill set requirement
> (Yet to start)
>
> 4) Identifying business partners
> (Got some leads in EE directory site, but do not know where to start)
>
>
> Pl do share whatever you think may help a new startup lab,so as to make
> it more effective
>
> Looking forward for your valuable guidance
>
> karthikeyan
>
I suspect that you're going about this backwards, or perhaps sideways.
If you first decide how many developers you need, then the capacity of
the processors, you will only succeed if you happen to luck out with a
set of tasks that fit that size team and that size processor.
I think that you'd be much better off having someone come in (a
permanent or temporary business partner, perhaps) to look at what you
want to do, translate your requirements into something that a
development team might like to see, then make some educated guesses
about the size of said development team and the requirements on the
processor.
I notice that you seem to want to start up-front with a processor
choice. If I were in charge of this effort I would make some general
decisions about the capabilities required from the processor and deduce
the processor size, then get input from the actual team on the processor
selection. I would do this because everyone has their own likes and
dislikes in processors, and getting to select the first processor will
make for a happier team. Part of the processor selection process should
be the selection of tool chain and RTOS (if necessary) -- tool,
processor and RTOS selection all interact somewhat, so it's best to make
the decisions at once.
I would make sure that each electrical and software engineer has their
own scope. In general the software guys just need scopes that will let
them see signals wiggle, and the electrical guys will need better scopes
to see _how_ the signals are wiggling. I'd get name-brand (Agilent,
Tek, LeCroy) scopes -- compare the cost of the engineer against the cost
of the scope; even if you're setting up in China it makes sense to speed
them on their way with good equipment.
The skill set should reflect what you're building. At a guess I'd say
you need some good basic digital, some embedded software, and probably
some general-purpose or power analog -- but as I said, that's just a guess.
I don't know what the expectation is for amenities in your area, but
where I come from you can have weight rooms and swimming pools and even
dancing girls -- but if management treats the engineers like dirt
they'll be unhappy and it'll show in their work. Conversely you can get
amazing things done in a poorly heated building with smells and noises
if the team really feels appreciated. Best, of course, is good
management, a warm room, _and_ dancers, but that's a rare work environment.
--
Tim Wescott
Wescott Design Services
http://www.wescottdesign.com
Posting from Google? See
http://cfaj.freeshell.org/google/