ITP – Be Here Now final project

I worked with Courtney Mitchell and Chris Egervary on Windowsill, our final project for the Be Here Now class. The idea was to design a flower pot that also contained a bird feeder and a web-enabled camera. The camera would take pictures either at scheduled times (sunrise, noon and sunset) or when a bird came to the bird feeder (triggering a PIR sensor & the camera shutter). The pictures would then be sent to a digital picture frame.

For the final, we were tasked with this design challenge:

Design an ambient device or service that can be used to display information from a network in real time.

The class had a group brainstorming session, where some of the generated ideas integrated pets. With all three of vaguely interested in pursuing this idea, we moved forward.

At our first meeting, we returned to the idea of Be Here Now and what it meant to us. While Foursquare and Twitter we mentioned often during class, the core idea that we kept returning to was one of mindfulness. Other similarly important concepts were being in touch with nature, simplicity, balance and mental health.

Our initial brainstorm ideas

We defined a core problem that we wanted to address in regards to the idea of Be Here Now was this:

People in the urban environment, overwhelmed by information flow, are losing touch with nature & simple pleasures of being alive.

As such, we defined our mission as:

Design a practical way of reintegrating a mindfulness of nature back into our busy lives.

Our target user was someone who works in an office and gets caught up in their Twitter feed. We imagined our service to provide a polite interruption in their day, one that gave them a view of the natural world outside their information flow. It assumes that they will take the digital picture frame to their office cubicle and will have it on their desk. This is the logo we went with:

On the technology front, we imagined this flow:

We began ordering parts, beginning with a waterproof webcam we found online. It also had a library of Arduino code. We also ordered a wi-fi shield, but realized the camera also required an SD card. When discussing this with Tom Igoe, he suggested we just use their beta version of the official Arduino wi-fi shield, which has a built-in micro-SD slot.

Having all been through the Intro to Physical Computing class, we assumed that we were not doing anything that cutting edge. We assumed that the Arduino could handle the tasks of taking a picture by set time, taking a picture by activation of a PIR sensor and transmitting the image.

testing the PIR sensor

Photos of Courtney and I testing the camera and the code for storing it on the SD card.

But in the end, we discovered that the Arduino technology is not capable of connecting all this functionality. Courtney took the lead on the coding front, in large part because we also needed to write a script to send the picture from the Arduino to a server. Then the server would send an e-mail with the photo attachment to the picture frame. The Arduino cannot handle the e-mail with the attachment itself. She was able to set up the server thanks to her experience from her Dynamic Web Development class. She also sorted out the requisite PHP script that moved the picture to the picture frame.

But the problem became the HTTP connection between the Arduino and the server. The 8-bit processor could not send the picture quickly enough. As a result, the server would time out and close the connection. This was the result each time, regardless of the fact that we set the camera resolution to the lowest possible setting (160×120 pixels) to get the smallest possible image size. Despite hours and hours of Tom’s very generous assistance, the experiment failed and we ran out of time.

waiting through one of many attempts to connect to the server over the wi-fi shield

Nonetheless, we were able to finish designing the physical device. With a flower bed, copper pipe, rubber tubing, a glass bottle and a few pieces of wood, we made the home for both flowers and the bird feeder. We liked the simple DIY look of the glass and copper. We also thought it went well with the wood. We filled the bottle with bird seed and designed a stand that held it upside-down. More seed would flow out as needed. Directly above a wood stick was the PIR sensor, aiming straight down to catch any movement in the limited area near the bird seed. The sensor was below the wood roof and the cables were wrapped in rubber to make them water-proof.

Chris glues together the pieces.

Filling the bottle with bird seed.

The wires for the camera and PIR sensor all were fed below the soil and came out a hole we drilled in the back corner of the flower bed.

While we were not able to get Windowill working the way we had hoped, we learned a lot along the way. As a team, we were happy with the product we designed for the assignment in line with the project brief. In pushing the Arduino as far as we did, we were able to better understand the limits of that technology. It was able to support most of the pieces of the puzzle (taking a picture, saving it to the SD card, making a connection with the server over HTTP). And Courtney was able to get the server and picture frame synced. We were just never able to get all the pieces working as a whole.

In the end, using a mobile phone was one approach we could have taken. Another is to have the picture go from Arduino into Processing and then into e-mail to the picture frame. But this of course means that a computer would need to be attached to the picture frame as well. All very good lessons to learn and things to build upon next year at ITP!