day2day: tokyo

Date: 2015
Location: Tokyo
Tools: Python, UR5 Robot, Processing, Arduino, Autodesk Vectorize It
Client: Autodesk Gallery
Description: A generation ago, taking, printing, and archiving photos was time consuming and expensive, reserved for momentous occasions. Today, we constantly upload smartphone photos to an ever-expanding web of social media. They appear momentarily on our feeds and then disappear.

We see a building, a pair of shoes, or a face that interests us and we snap it and upload it. Through automated geotagging, these moments in our lives aggregate into a shared social desire of a specific time and place.

day2day is a robotic installation that celebrates time and place through the painted transcription of this collective yet disparate narrative. It searches a chosen location and selects the most-liked photos, then custom scripts convert the linework into instructions for a robot to paint. Over a two-week period, the robot will produce a scroll of 250 individual scenes: a hard copy of our shared lives as seen through social media.

The installation consists of a universal robot (UR5) equipped with a custom end effector; a website that serves the robot files via Instagram and displays the final results; python scripts that create the robot’s toolpaths; and a custom table with an automated paper roll. These components document a geolocation's Instagram feed in real time using an ink brush.

Credits: Matsys (Andrew Kudless with Alan Cation and Clayton Muhleman), Web development: Richa Agarwal