Unless your products are Mobius strip connected Klein bottles you generally don't need to have integration between the mechanical and the board layout engineers. As soon as you do that you are forcing someone to use a tool they probably don't like or don't want to waste exorbitant dollars on. This can EASILY cut out the very person you should be using because of his/her skilz and technical panache.
The normal way you do this so everyone can do their jobs in a friendly professional manner void of extra stress not a single soul needs is to sit down with the EE and (1) go over what the machine is and how exactly it operates.
After the basic needs operation is understood by the EE go (2) next over any potential additions that could be needed if system issues present themselves. "We could need heat tracing on these 3 pipes but we don't think so." "We might need to control an additional valve for the hydrazine."
(3) Ask the EE if there are any features or additions that might make sense. This could provide some serious product benefits for amazingly little expense. Improvements that could easily be completely out of reach logically or cost-wise if they needed to be implemented later. Give these ideas
serious thought. Avoid tossing them in this meeting. Discuss them with others, especially the sales/field people. Numerous times I've shocked the MEs with possibilities they had no idea could be added without much or any expense.
(4) At this point go over what's needed for a user interface in general terms. "Buttons for flush, regurgitate, swill dump, etc." "A knob to control churning and two temperatures."
If you have a specific interface look that is needed for a proper product line feng shui then provide examples to the EE. This will be the guide. If not let the EE ponder it for a day and get back. The EE will be the most UP on the latest way to do Human Machine Interfaces (HMIs) and can give you some examples and pros and cons.
More than a couple of switches and buttons will quickly lead to maintenance and needless wiring expense. Head for a graphical touch HMI if the product exceeds a very few buttons and knobs (steam controls). Avoid steam controls immediately if different product models would result in a single change to them OR if the product is poorly defined at the moment and may need addition controls.
But I digress.
(5) Ask what kind of space the EE needs for the controller or show a proposed area with general dimensional limitations and talk about it. If it's immediately tight the EE will need to ruminate on the situation. Could be a multi board design will be needed. Perhaps the system requirements speak to a distributed control system so the controller can use a couple of interstitial spaces effectively. Perhaps he can educate you on reliable controller needs like not roasting it. Or, putting it somewhere dollar connectors can used instead of $25 dollar connectors.
On a different plane if a product needs to be really crammed it may make more sense for the EE to do the packaging with much communication and approval and mods coming from the ME side. An example might be a hearing aide or a cell phone.
In the current, completely insane semiconductor supply chain situation, I can only recommend you don't bother thinking "real small" unless you're okay with year long production blocks. Things like processors are often single sourced and I've seen ones that were $6 now being over $100. You can comb the planet for a common commodity part of two years ago, end up at a grey market vendor that demands 3X what the part was priced at paid up-front. A week later they inform you the price has gone up 25%, pay another 25% or they'll refund you. Now you're a week later, you have to pay. Requiring smaller parts can add to this dilemma.
Communicate. Wait as long as possible before locking the controller-space. Communicate.
Keith Cress
kcress -
