Building and Materials

After collecting resources from the workshop room, including the wooden hand, wires to move it, servos and pulleys I started building the final product. Working off the feedback from the prototype the goal is to make a more sturdy hand where the fingers can move freely and be positioned in any way I wish. The wooden hand has stiffer joints which enable the fingers to stay in almost any position which should come in handy for that. With all the wires attached to the fingers however, it is clear that they will not pull themselves back up like the elasticity that was in the cardboard hands so wires will also be needed at the back to pull the fingers back up. Additionally wires will be attached to the wrist and back to the servos to move the hand, and need space to rotate and move around in order to properly pull the strings.

As you can see below, I made a huge mess with all the materials and building, where most of which was done by trying to solve a problem using trial and error. Building using odds and ends I found at home, I used a plastic container for the arm which turned out to be quite sturdy, but needed to drill holes into the top and sides to attach the servos in so they could rotate it. A servo was attached to the wrist and the hand was pulled off the of the arm it was attached to in order to attach the wrist servo to the base of the hand.

Putting it together

The foam base from the prototype worked very well and the servo was placed inside the plastic container for the next stage where all servos were in place. Another box was used as the container for all the wires and Arduino back end to put it all together and out of sight. With the physical components set up, the next stage was to connect all the wires on the hand to the servos, code the Arduino components then attach it all to the box and battery to make sure the hand stays where it is while getting powered to move and perform gestures.

With everything in place the wires were loose to not stretch themselves out of place before the exhibition while getting all other parts in place. Glue and lots of tape was used to keep the hand in place, all weight resting on the arm servo as the hand will have to move independently to wave, thus not resting its weight on the base of the arm at all. The two distance sensors were placed outside the box facing forwards, offering a simple method to see from which direction people would approach the concept, then the arm would rotate to face the side from which they approach. This was done easily by comparing the distances at each and then using the distance of the side which was closer, and then the action relevant to the distance detected would be performed.

After all the time and effort payed off, the final product was finished and worked decently well, however the battery runs out very quick and when I attached the hand it turned out to be very difficult for the servos in the wrist and arm to rotate. I am quite happy with the look of the final product and believe that it works quite well for the intended concept, and if it wasn't too heavy it could have been a successful design.

Additional feature

This week I added the final interaction to the prototype. Based on the feedback I had gotten on the prototype demonstration I decided to add a way for the users to delete the message instead of sending it, if that is something they wish. During the prototype demo I also got confirmed that the throwing interaction was an appropriate metaphor, based on this I decided that the delete metaphor should be the opposite, namely drop the ball instead of throwing it upwards. This, however, turned a little more difficult than anticipated since it required the accelerometer installed to know the difference between a throw and a drop. By studying the output values registered when performing these various task, it became apparent that they behaved differently from each other. The throw slowed down at the top of the throw before it falls down again, while a drop has only a quick stop of the acceleration. In the end, I managed to make it work.

Because of the time restriction and the restriction I won't be able to conduct a user testing to confirm if this last interaction is the best way to delete the message. A different metaphor might be even better than the one I have chosen but hopefully it will be easy for the user to understand and interact with.

Next week

Since the prototype is almost finished, I will start to make the website for the exhibition. I have already some ideas of the design that I want to create for the website and hopefully it will turn out as good as I have imaged. Building a website is just fun in my opinion so this is something that I look forward to.

I spent most of my time last week working on the next prototype and my annotated portfolio.

Studio & Workshop

In the studio, we had our regular stand-up with this week's focus on having a one-line pitch for our concept, show what we have been working on, what our priorities are for finishing the project for the exhibition and questions regarding the portfolio. Although we have slight variations of pitching the concept and thoughts of the ideal product, given that we are still exploring different aspects, our current one-liner is "E-mories, a distraction-free physical platform to remotely share personal emotions with close friends and family". Moreover, for my progress, I showed the state of the ball, which at that point in time was the new ball with a bend sensor attached with silicone to the inside. As for priorities finishing the project, we had already fixed it so that all devices could communicate over the server, so we mostly just had to continue focusing on making sure our individual prototypes works and to conduct user testing for the last prototype.

Prototype

As for the prototype, I found a nice transparent ball at K-mart which I could use. The ball had a nice pattern to it which I believed could reflect the colours in a neat way and it also contained some glitter water inside. At first, I didn't think much of the glitter water as I mostly wanted to use the ball. However, looking back at our additional features suggested in the proposal, one of them was to add water to the E-mories device. Given that I am building this prototype to test material and how to make the device more of a personal artefact, I decided to test how adding glitter water could make for a unique look and feel and test whether it made the device feel more personal.

I drained the water from the ball and started to place the various Arduino components and sensors inside, making sure they were all fully covered in silicone to avoid any water touching the electronics. I covered all electronics I could in clear plastic wrap and black tape before I covered them in silicone, to better protect the electronics. I let the silicone dry before I carefully tried to put some water inside to see if it was still leaking. Three times I had to put more silicone in to stop the ball from leaking, which was strange as I by the end had covered the entire bottom half of the ball with silicone.

When I finally had made sure it did not leak, I tested to see if everything still worked, which it did, but the accelerometer has ever since testing it for the first time seemed quite unreliable as it seems to mix angles. As for the working parts, they can be seen in the images and video below. At this stage, everything from recording to picking a colour, sending data and getting an incoming message notification worked as intended. However, when I picked the prototype up the next day, the bend sensor values were all over the place, which made nothing work. I inspected the ball for water leakage, but there was none. I knew from when I received the bend sensor that the connection was somewhat loose which I had taped earlier to avoid these issues and later put silicone on top of to hold in place. Despite this, I seem to have some issues with the bend sensor. Having tried to fix the issue for a couple of hours, I decided to drain the ball of water in case that had any effects on it. So, I'm going to let it dry off before trying again. If not, I might have to stimulate the squeeze interaction as I would have to pull everything apart to access the bend sensor and fix it at this point, basically meaning I would have to purchase a new ball and start over from scratch.

Web Portfolio

As for the rest of the week, not counting the time I've spent working on my thesis, getting ready for the prototype demonstration there, I spent working on the portfolio. My current progress can be found here: portfolio

Most of the time I've spent working on the portfolio has gone to rewrite and condense what we have already written about in the proposal, journal and first prototype delivery. Re-writing this content feels rather repetitive and as a result, made my motivation take a hard hit. I'm still struggling with motivation in both courses as there is a lot of repetitive work and every day feels the same, not being able to have a social life for the entire semester. Still, I believe I'm on track for the exhibit and portfolio in PhysComp, while I still need to catch up on the thesis as PhysComp requires most of my time throughout each week.

Work Done

Spud

I have been working on implementing the ultrasonic sensor and the voice recognition and have met with difficulty in making them work together without confusing the Spud program. As the ultrasonic sensor and the voice recognition software are always on, I have to decide on giving one of them the priority of making an action. Otherwise, as I'm experiencing now, when Spud detects someone nearby and listens to a voice command, it does the actions one after another without stopping. Therefore, I have decided to make voice commands the priority. Also, I have to think of a way to loop the angry expression and the stop motion where Spud shakes its head as long as it is detecting that someone is at that distance. Otherwise it would be weird if Spud just goes back to a neutral position after a while even with the person still standing there not taking the hint when it is supposed to do try and make that person to go away.

On the mean time, I did testing on the interactions of Spud with participants. I tested whether the participants understood the meaning of all the different movement and how they felt for each of them. A few commented that the movement was too slow and the dancing trick looked like Spud is angry instead of being playful. In response to this, I will increase the speed and try to make Spud look a bit more silly instead of angry.

Website

I first planned how my portfolio will look like and have now completed the overall frame of the website and chose its colour scheme. I also drew some illustrations and icons of Spud so I can add them into the website.

I will complete all the functions and features of Spud as soon as possible and conduct a final user testing observation and interview while adding details and information into the website.