The career fairs at Georgia Tech are pretty cutthroat. The main fair during the Fall semester is attended by over 200 companies and a large portion of the twenty thousand students who attend GT. Safe to say, standing out is difficult. I didn't have much luck this past year, and I kind of had to go with anyone who would offer me a position, so I'm looking to step my game up this round. I've decided to create an electronic business card.

market research / a summary

Plenty of people have done PCB business cards in the past, so there's even some challenge standing out from the prior art. It seems the most popular designs merely blink some LEDs activated by a button powered by a coin cell. Creative versions will use a capacitive touchpad and a small microcontroller. Coin cell batteries use bulky holders, which make this type quite thick and thus less likely to be stored in a recruiter's wallet. These are also quite uninteresting, you don't have to take a circuits class to light up an LED, so these type don't do much to broadcast much of a skill set.

On the contrary, there are some very well done cards. This one, by Limpkin, uses two 1.0mm PCB's in a sandwich to allow the card to slot into a USB port for power. I have seen 2.0mm boards before to allow powering via USB, but the sandwich allows the components to be flush with the surface of the card, so the card is easily walletable.

Others use the USB port for power and data, the most common of these types simply enumerate themselves as a USB mass storage device with a resume already on the drive. These are pretty uninteresting to me, though from a marketing standpoint they are optimal. Assuming the card is somewhat thin there's a good chance that anyone that is given one will keep it around to use as a USB drive, increasing the exposure to your name and contact information. This is something I wanted to encourage with my ideas.

There are more interesting ones in this space. Frank Zhao's card emulates a USB keyboard, and waits for the user to open notepad and toggle caps-lock three times, after which his resume is typed out by the card into the open notepad. Ch00ftech does something similar but emulates a USB absolute-positioned mouse instead (like a graphics tablet) to draw his logo into a paint program. This one by Ramiro Veredas is just a mass storage device, however instead of mounting his SOIC package on top of the board, he mounted it through the board, reducing the overall thickness of the board considerably.

The last one that caught my eye was this persistence of vision card. It's very bulky, so it's not going to fit into any wallets, but the design is incredibly interesting and well done. It also does something interesting immediately after you hand it to them. No need to make them wait to get back to a computer at the end of the day (when they're probably tired from talking to 100s of other eager students throughout the day). When you hand someone this, you can go "here, watch" instead of "yeah! well, when you plug it into your computer... just trust me on this."

some planning

So looking into all of this, I needed to set some goals. Here they are.

  1. It must be walletable. -- There's no point in spending so much time on something that whoever you give it to will just throw it in their bag and, if you're lucky, they'll place on a shelf somewhere. If it's compact enough to hang out in someone's pocket, you increase the chances that they take it out and show it to other people, some of whom might be their coworkers at your dream company.

  2. It should to something that's immediately interesting. -- Immediacy of the cool is the key here. Like before, it's cool if it does something, but it's an order of magnitude cooler if the person you're showing it to doesn't have to trust you that it does something interesting.

  3. It should have some utility. -- You want to give whoever you're giving the card to a reason to keep it in their wallet. The longer the card is in their wallet, the longer your name is in their wallet, the more likely it is that they remember your name the next time they're looking to fill a position. -- If it is a USB drive, put some extra space on the drive so it can be used to transfer files. Even if the card lights up, make it configurable.

  4. It should be hackable. -- This is potentially just an increase to the utility. If your card is hackable, then there's a potential that whoever you give the card to wants to change what it does, or shows it to someone who wants to. There should be debug pads and programming ports readily available.

  5. Make it open source. -- You can't forget that the card is a simply a vehicle for your expertise. Having a card that works is impressive, but if an engineer at the company can look up your board design and see how much work you put into routing the PCB or how clean, commented and well-structured your code is, then you just became that much more valuable in their eyes. In this vein, make sure the code and PCBs are clean and thoughtfully made, make it your best work.

the idea

So, I really like the persistence of vision display. It has the immediacy factor that I want. The only problem I have with it is the utility. It's really not that interesting to have a PCB that says some random guy's name when you wave it in the air. If I'm waiving my hands around like a maniac it better be for good reason. I also want to increase the hackability, theres not much more you can do with some LEDs and a tilt switch.

To fix the utility issue, I'm going to include a light sensor of some sort (photoresistor/photodiode/phototransistor/ambient light sensor IC) to allow for asynchronous serial communications over light (think, UART over light, or a series of flashes), most likely originating from a webpage. This will allow for the message that the card displays when shaken to be changed using a web browser on a phone or computer. Hopefully this will convince the user to keep the card around indefinitely, so they can display different messages for different occasions on a moment's notice, such as "Happy Birthday" or "Congratulations, you're hired Matt!".

I also want to go with a digital accelerometer instead of a tilt switch, since it will make the card more adaptable if the user wishes to modify it's behavior by hacking it. It could be turned into an adjustable brightness night-light, or a bubble level, or a number of things. It will hopefully also allow for some data processing to improve the quality of the display.

The thinness is going to be the biggest challenge. I'm not 100% sure how I'm going to solve this issue at the moment. For now, I'm going to start on a prototype (I want to have something working by the career fair, even if they're not good enough to give out by then) while keeping the thickness of the finished product in the back of my head.