Hackathon: "Physical interaction"14 september 2017
Planning and startup
Team two was off to a slow start because the team was in disarray due to an unexpected loss and subsequent addition of a new team member. But the addition of Johan added much needed inspiration and the team got of to a late but quick start when the team gathered on Wednesday for a brainstorming. We decided on making some kind of physical interaction with an online gaming experience. Since no-one came up with a really good idea we could unite about we left with a set goal that we should each think about it overnight. A couple of hours later Åman exited the team as abruptly as he entered, but not before he purchased the necessary equipment during the evening.
Late in the evening Mats built a frame for the range sensors with some leftover pieces of wood from an old sauna.
The remnants of the team gathered at Toftagården in the early hours of Thursday september eight in a room called Lilla salen(Small ward). Luckily the room did not live up to its name since it actually was quite spacious. During the first hour of the day we focused on getting the Arduino micro-controller working and setting up the development tools which was easier than expected. Meanwhile we discussed what we should do with the Arduino. We settled on making an physical interaction with the Triss product using distance sensors connected to the Arduino. Mats and Kristofer worked on getting the range sensors working with the Arduino and getting the measured values to the computer via serial communication. Meanwhile Bodil worked on the ability to interact with the scratch-code in the triss from an external data source. Philip and Andreas began writing their own implementation of the Triss as a backup if the original triss code was deemed to complicated to interact with.
Since the sensors did only have a 15 degree measuring angle we could not use the wooden frame since the virtual scratch-area would have been less than a square of 8 by 8 centimeters. So instead we mounted the sensors opposite each other so we could control x-axis and y-axis with a hand on each side of the controller.
Day one evening
Kristofer got the code working to send data from the Arduino via node.js to a webpage using socket.io He also added a LED to each range sensor so we could visualise the interaction with the range sensor. Bodil got interaction with the triss working and the Philip-Andreas triss was also up and running. There was a serious incident with a beer bottle and the Arduino late in the evening, but luck was with us and disaster was avoided in the last micro-second.
Kristofer continued with his audio-visualisation by adding an audio component to the range sensors for improved WOW-effect. Many R2-D2-like sounds were heard in the room while he tweaked the audio with relentless vigour. There were mixed feelings from some other teams and there was actually a WTF exclaimed by a competitor. Bodil worked on getting data from Kristofers socket server with mixed results. The communication was not as stable as we would have liked.
One of the last problems we had was getting values in the correct range for matching the triss scratch-area. The range sensors didn´t seem to be calibrated so first we had to compensate for the differences. The sensors sometimes delivered values that were not realistic so we had to filter those out to. When we got all the values for the original triss, we continued to get it working for the 30, 40, 90, dubbel and mini variations of triss with the possibility for more variations with four lines of code each. In the nick of time we also got the scratch-tool working by some heroic last effort coding by Bodil.
Watch a Demo
It was quite fun to build something physical and in the end watch the result as a scratching of the Triss online.