Talk about preference. I asked a question this month that could have gotten just as many responses as people responding. The question was "When you sit down to start a brand new design, what's the first thing you do?"
Of course, the answer has a lot to do with what you are designing and what's been done before. For example, if it's a portable device, I may start to think about the power requirements. If it's a medical device, I'd probably start by making a list of all the standards bodies I need to be in touch with and all the regulations I need to comply with. There's also the question of the geography in which the design will be deployed. Hence, there are likely even more answers than people answering the question.
Regardless, I thought the question would bring some intriguing responses from our readers, and I was right. For example, Eduardo Viramontes said:
I talk to colleagues and team members about the project to get their input and ideas. Some people will focus on very fine details like, "remember to put a pull-down resistor in such and such pin." Others will likely have more high-level ideas like, "you should put a bunch of LEDs here to display a smiley face." These types of unedited ideas, unburdened by the development of the project, will generally only come up in the beginning, when there's no bias about the specifics of the project. And they can save a lot of time down the road.
Emilijan Iljoski said, "I first review my previous designs and analyze the existing designs of others that may be related to the subject matter. De facto, to design something innovative, you must know what has been already done."
That makes sense to me. Basically, that's saying that you don't want to repeat the mistakes of others. Rather, you want to learn from them.
Soumanou Eusebio said:
I like to start with a literature review of existing design, then hold a brainstorming session with my group. We put down all the ideas; regardless of whether we think they are good ideas or bad ideas at this point. Finally, I will narrow the ideas down to three concepts that I will study, test, and compare. From there, I get my final concept for prototyping.
Edna Elizabeth Ramos Cienfuegos said:
To begin my design, I have to know the "voice" of the customer, which I translate into specifications. At the beginning of the design project, it's important to have targets regarding cost, time, quality, and the segment of the market that your product (or service) will be deployed. For example, in France they have "cahier des charges." This is similar to the idea of customer-to-customer, or C2C in the US. It's also related to Advanced Product Quality Planning (APQP) or Concurrent Product and Process Design (CPPD), or any process that requires the understanding of the inputs to trace the boundaries of the design.
And finally, from our own Jon Titus: "I make sure that I have clear objectives and requirements written and agreed upon by all concerned parties. You can't start a project without a clear 'map.'"
Readers, where do you start? Tell us in the comment section below.
If we are talking about consumer products and ones that will be used by individuals, we are working in a direction somewhat determined by codes and standards; i.e. UL 858 for electric cooking products, Z21.1 for gas cooking products. Jigs, fixtures, dies, tooling, robotics etc give a little room for creativity because they will be used in-house relative to the manufacturing floor. In looking at products that will ultimately be used by consumers:
Have a complete definition of the product needed and the scope of the overall project. What do the customers want and how have those needs been defined by marketing. If in doubt, obtain the necessary clarifications up front. Don't wait until the first design review to obtain additional understanding of what's needed.
Obtain an absolutely clear understanding of time-lines for the project, including design guidance, design confirmation, pre-pilot, pilot and production dates.
Obtain the LATEST codes and standards to which the product will be tested. (This is critical and includes all local codes and any national codes.)
Discuss the product and project with manufacturing to make sure there are no issues relative to the overall project. (Don't throw the design over the wall upon completion and expect manufacturing to make it. Be up front and make the design/manufacturing phase a joint venture—no surprises! )
If components now used in existing products can be "designed" into the "new "product, by all means do so thus eliminating an additional item to be purchased and manipulated by assembly.
Make plans to communicate with manufacturing, marketing and upper management on a weekly basis the status of your project. Set a time and day each week for a brief status meeting.
If needed, have legal determine if infringement in any area might occur.
If needed, research in-house and in the literature designs of a similar nature. Start with in-house.
Everything that I design is a "custom" design, in that it is created in response to a request for a product. OUr best customers arrive with a list of what the designed product must accomplish, and how fast it must accomplish the task. Some customers arrive with a list of what the system must do and how it must do it. They may, or not, understand what they need. So the first step is always finding out what the customer needs.
OF course, I am often at a good advantage in that area, because to produce a price quote for a product, it must be fairly well defined, and so going over the technical proposal usually is a very good first step. But we always meet with the customer to discuss just what the product will do to benefit them, since that is sort of mandatory information.
I take as a given that I will be asking others for input and trying hard to learn what else of the same sort has been done.
However, the first thing I do is I start imagining. First I imagine the customer... and I may not know MUCH about that customer to start with, and I will have to learn but I have to start by imagining (and correcting the image as I go) the customer. Their experience is primal in this process.
Second I imagine a device to do what the customer needs and wants.
Third I imagine the device I have imagined working. It works to do what my "customer" wants in some fashion and I have to imagine the way it works.
Starting vague and refining and restricting the design until I have something useful or nothing is left and I decide it can't be done.
warren- got a real chuckle out of "chief technologist = old engineer". Thats the way it works around here. My boss, the "chief engineer", is a title (he is a graduate civil engineer with zero experience at anything). If asked any technical question, his standard response is always "go ask Al"!
In my experience, the customer usually does not know exactly what he wants other than he wants it cheap, fast, and good. At this point I usually explain that there is a natural law that he can have only two of the three. Something like Boyle's Law, only different. Sitting down with them and trying to see their vision is the first step. Then look at their specifications and modify it as needed. Next I simply stare outside, with a cup of coffee and a cigarette in hand, and mull it over in my mind. Many times a great solution pops in my mind while sleeping. I then sketch it out on graph paper (always graph paper for some reason!). Then at work I start filling in the blanks with all the "hows" to accomplish the task. After that the computer work starts, then debugging the new hardware/software. Midstream in this I kick myself for dropping out of medical school.
My first cut is seldom correct, chop up the prototype, mangle the microcode, and viola- a work of art that performs (usually) like a Michelangelo but built internally more like a Picasso!
I think every company should have their own Design Requirements Manual that specifies engineering and hardware requirements for the typical product design. This manual would also include theory that explains why (historically), some aspects of every design are similar or the same. For example, certain materials are FDA-approved; some materials have proven to be best for marine environments or have been thoroughly tested for UV resistance.
This type of manual is especially handy when tribal knowledge gained from years of research is not readily accessible. So, assuming this DRM has already been created, the first thing I would do is read it. If it doesn't exist, create one, yourself after obtaining/learning what needs to be in it. If you are new to the company, this may take a while and most certainly will be a result of input from many people within your organization. Change to the document will inevitably occur over time, but it's engineering fundamentals should not.
I wholeheartedly agree, assuming the ultimate goal is to sell a product and get $$$ from customers for doing this. To understand the problem to solve you have to understand customer needs. And what's the first step in problem solving? Making sure you understand the problem. A good marketing organization will be in touch with this (sorry Battar, I disagree with your first point). And on a larger scale the "customer" could be in fact a "market" (a market is a group of customers with a particular set of needs). Needs become requirements, then research for solutions, which becomes a block diagram, a design proposal, then finally a design, all while verifying with the customer (or market) it will satisfy the need. Engineers often find the customer knows they need something but can't put their finger on it so they need a little help (again, a good marketing organization...). And yes I do agree with Battar, at some point you really do need to stop designing.
When the project starts is different for each of us. Some of us may have a written specification...... OK, a startling few of us will be afforded that luxury, but if we are designing for an OEM they probably have a written specification. If I have a written specification the first process is to determine if the specification makes sense, and do I understand what the customer wants. I have found that the User Interface is often confusing in customer specifications, and what seems like a good idea on paper becomes very confusing when people really start pressing buttons.
Most of my projects today are new technology and the specification is conveyed over a phone call, "Hey, is there any way we could do this?" For those projects I search for any prior work that can be helpful and also signal any patent problems before the development goes any further.
Somewhere in between are the projects that turn into real products. There is no written specification, but it's an appliance that is familiar enough to design into a prototype that can be handled by everyone involved in the development so that the minute details of the design can be completed. Yes, a written specification is better, but getting a prototype into the hands of people really brings the project to life.
It might not be simple, but incorporating the voice of the customer in critical and should be the responsibility, in part, of the development team to address in some fashion. Maybe they're not asking the right questions or even the right people. But developing in a vaccuum has been the death knell for many a good product.
Great point, naperlou. Before I read your comment, I was about to offer kudos to those who voiced the needs of the customer. But your comment reminds me of all the times I've interviewed engineers who said that their customers often didn't know what they wanted. Sometimes, it not as simple as it first appears.
It won't be too much longer and hardware design, as we used to know it, will be remembered alongside the slide rule and the Karnaugh map. You will need to move beyond those familiar bits and bytes into the new world of software centric design.
Focus on Fundamentals consists of 45-minute on-line classes that cover a host of technologies. You learn without leaving the comfort of your desk. All classes are taught by subject-matter experts and all are archived. So if you can't attend live, attend at your convenience.