I'm working on a new project right now, an installation piece for my living room that I'm calling the Bus Stop Sign. The plan is to use the public data feed from Boston's MBTA to create a permanent, wall mounted version of what's essentially the NextBus App, displaying the time until the next bus arrives at the two stops outside my house.
It's a piece of what I call "Single Purpose Dedicated Hardware." Of course, you could do the same thing with an app, and there already is one, but I like dedicated hardware for a number of reasons. First the UI is simple and will need no buttons. Second, it's always on and ready to go. Third, it will look fantastic, in a way that no phone screen or website ever could.
The mockups below show my initial concept (2 bus routes, with 2 directions each, running to a possible 3 destinations) and read on for more detail.
I'd like to take a moment to talk about my process and approach to making something like this. The framework I use is a phased approach, but I'll heavily blur the lines between phases as needed, and this project has less technical risk than others I've taken on, so it compresses some steps. I'm on the fence about commercializing this, so I'm unlikely to take it past the Looks Like/Works Like stage described below.
Concept Stage and Prototype Stage:
These stages (combined together on this project) are about the forming the idea and making a mockup as soon as possible. For me, that meant writing a Python script to process the data feed to prove the concept and some industrial design. On a more technical project, I'd also include a hacked together version of the hardware in this stage.
Looks Like / Works Like
This stage requires work across 2-3 major disciplines. I need to design and pick out exact components for the circuit board and 7 segment LCD displays and I need to get the industrial design and appearance of the sign squared away enough to nail the mechanical design and mounting of the components.
In terms of circuit boards, my plan is to create "breakout boards" for each of the 7 segment displays and to harness them together using an IDC Ribbon cable. The harness will tie them back to a segment driver, which will talk over I2C to a BeagleBone or RasPi which will act as the brains.
I have circuit boards on order and components in house for hand soldering, so this portion of the design is a bit further along than others.
For the industrial design and mechanical assembly, I'll probably opt for a laser cut housing, although I haven't settled on a method for applying my graphics. (Silkscreen? Laser etch? CNC Router? Spray paint stencil? Hand paint? I don't know yet.)
It's unlikely I'll push the project this far, but this stage would be about figuring out which sources and vendors to use for the parts from the previous stage and organizing a Bill of Materials that can be taken to a vendor and quoted.
This stage is about making tweaks and improvements using vendor and manufacturer feedback, making detailed assembly directions, test jigs, QA fixtures, etc... all in support of higher volume work.
This stage, I'd hope would be self evident, and is about making things in volume.
Anyway, as WrightGrid keeps me quite busy, my plan is to continue sharing and developing these projects as I make progress. Commercialization really means pushing through the Quoteable and Preproduction stages, which requires a LOT of extra attention to detail, plus all the collateral support for a marketing / Kickstarter campaign. I of course, don't think I can ever stop making things, and a project like this which involes high level software for the web based stuff, low level software for the I2C driver stuff, schematic capture, circuit board layout, graphical design, industrial design, and mechanical assembly, well... it keeps me sharp ;-).
aka THE Awkward Engineer