When you're short on time what are some things that help you save time on projects?

Keep it simple. Simpler systems have less latitude for mistakes to occur. They are easier to debug, and can be easier to evolve into more complex systems in the future.

I find the best work gets done when you are low on time and resources. The two Mars rovers (6 months to design and build, designed life 90 days, one is still running at nearly 3000 days) and ARM processor (no money, 2 people, now worth billions) are prime examples of this.

Make sure people know you are in a rush. Otherwise they will pester you with "Oh can you just do this? It won't take 5 minutes". Those 5 minutes add up and often become hours, plus the time you need to "get back into" your project.

Do it NOW. Forward planning is great if you have the luxury, but it can also give the illusion that you have more time than you really do. I'm terrible for putting things off, so I get it done and out of the way. Use the extra time for another iteration loop if need be.

Forget formal planning methodologies. Sure, elicit the core requirements, but filter out the ones that really can wait. Quite often this means building in an inherent flexibility into the architecture which pays dividends later. The Mars rovers FPGAs weren't programmed when the rocket lifted off - the software and firmware was nowhere near ready. As a result they ended up with a system that was resilient against faults, and which has been remotely upgraded several times.

Know your resources. What do you have that can give you a step up? Chunks of software? Dev boards and modules? Prototypes that you made for something years ago which can be adapted and hacked. Having a huge amount of "stuff" acquired over the years really helps, whether it is a stack of passives, or general purpose semiconductors, plus various converters, op-amps, embedded micros etc. Making your own PCBs has become unpopular in recent years, unsurprising given part density, multi-layer boards and so on. But having the ability to produce relatively simple PCBs (e.g. break-out boards) on site can save a couple of days. Knowing someone with these skills (or for that matter other technical skills which you can call on) is worth its weight in gold.

I totally agree with the work at home comment. This isn't always possible, but if you can work somewhere where you are undisturbed my managers, meetings, phone calls and other peoples' demands on your time, then you can work twice as fast.

Oh yeah, and living 30 minutes away from Farnell's head office and trade counter really helps!

Keep it simple. Simpler systems have less latitude for mistakes to occur. They are easier to debug, and can be easier to evolve into more complex systems in the future.

I find the best work gets done when you are low on time and resources. The two Mars rovers (6 months to design and build, designed life 90 days, one is still running at nearly 3000 days) and ARM processor (no money, 2 people, now worth billions) are prime examples of this.

Make sure people know you are in a rush. Otherwise they will pester you with "Oh can you just do this? It won't take 5 minutes". Those 5 minutes add up and often become hours, plus the time you need to "get back into" your project.

Do it NOW. Forward planning is great if you have the luxury, but it can also give the illusion that you have more time than you really do. I'm terrible for putting things off, so I get it done and out of the way. Use the extra time for another iteration loop if need be.

Forget formal planning methodologies. Sure, elicit the core requirements, but filter out the ones that really can wait. Quite often this means building in an inherent flexibility into the architecture which pays dividends later. The Mars rovers FPGAs weren't programmed when the rocket lifted off - the software and firmware was nowhere near ready. As a result they ended up with a system that was resilient against faults, and which has been remotely upgraded several times.

Know your resources. What do you have that can give you a step up? Chunks of software? Dev boards and modules? Prototypes that you made for something years ago which can be adapted and hacked. Having a huge amount of "stuff" acquired over the years really helps, whether it is a stack of passives, or general purpose semiconductors, plus various converters, op-amps, embedded micros etc. Making your own PCBs has become unpopular in recent years, unsurprising given part density, multi-layer boards and so on. But having the ability to produce relatively simple PCBs (e.g. break-out boards) on site can save a couple of days. Knowing someone with these skills (or for that matter other technical skills which you can call on) is worth its weight in gold.

I totally agree with the work at home comment. This isn't always possible, but if you can work somewhere where you are undisturbed my managers, meetings, phone calls and other peoples' demands on your time, then you can work twice as fast.

Oh yeah, and living 30 minutes away from Farnell's head office and trade counter really helps!