Knowing a bad idea is sometimes easy… check out this hellarious site: stupidbusinessideas.com.  What do these have in common? Simple: it is obvious that people are not willing to pay for it (or the market is VERY small).  But how can an engineer-type thinker discover if his idea (which isn’t so obvious) is a Bad (small) business idea?

I will outline 2 methods, which would ideally both be used.

The First Method P.O.D.C.
PODC is a qualitative tool that helps a person analyze if the 4 key components of a Business Idea make sense
P = People
O = Opportunity
D = Deal
C = Context
If you see green lights all down the PODC list, you are good to go to the next method (or take your chances if you wish).   Here is what Green Lights should look like.
P.  The people running the business and inventing the product(s) are experts in the relevant fields and have experience having done it before.
O. The opportunity (the problem you are solving/need you are satisfying) is large enough that a small market share is all that is needed for a viable business.  E.g. a 1% market penetration would lead to a $5MM+ business.
D. The deal proposed is not too small or too big and has milestones and success markers along the way before further investment is needed.  E.G> a small amount of $ investment $ is all that is needed to get to the first milestone.
C. The context (the world view) is fresh and new, not old and boring.. and the business does NOT rely on some obscure law or quick trend.  Competition may or may not be there, but this solution seems unique and could be a long-lasting business.

Did your PODC come out green?  Now the real test: will people pay?
To test this, Engineer-thinkers need to deploy a Minimum Viable Product (MVP).  This is VERY hard for Engineer-type thinkers, who want to solve the whole thing & get it working first.  WRONG.  You need to TEST to see if people will pay first.  
Some guidance may be found here: http://venturehacks.com/articles/minimum-viable-product  and http://venturehacks.com/articles/minimum-viable-product-examples and of course http://www.startuplessonslearned.com/

Here is my quick instruction manual to make an MVP.
1.) Find someone creative who will build you a 3-page website for cheap.  ($1000)
2.) Build page 1 that gives the pitch for the new product/service.  1-button: Order Now.
3.) Build page 2 that shows the price & looks like a shopping cart.  1-button: Proceed with Order
4.) One page 3, say: “Sorry, we are currently back-ordered.  You will be emailed when new inventory is available.  Thanks for your business.”
5.) Build some ads for your business idea on Google Adwords: http://adwords.google.com/
6.) MEASURE how much it costs to bring ppl to your site… how many are interested in Order Now… and how many “Proceed with Order” after seeing the price.
7.) Adjust parts 3-6… and see if you can make MONEY with the business.

If you can sell for more than 4x the cost (where cost is product cost + shipping + marketing cost)  you are in good shape.

If not, start over with PODC and see what was wrong.

Best of luck!

Now, get some good ideas!