The final project is intended to let us expand on a prototyping technique or techniques that we learned previously in the course. The goal is to build a prototype for the purpose of exploring the desirability or feasibility or usability of a design.

Concept

Last quarter, we had a trash project which is aimed to encourage people to sort the trash. At that time, Donglu had the idea to tell people the type of waste by scanning it, and Liusu had the idea of an educational robot that can teach people how to sort their waste by scanning the items. We decided to team up and prototype a robotic trash can — Wall-E. The goal of the project is to help people identify recyclable items and increase their awareness of recycling their waste.

The functionalities of this robot include: mobility, scanning, and speaking. The robot is able to move around automatically and also remotely controllable. When it meets a user who is about to throw some waste, it will ask if the user would like to recycle the waste. If the person would, she or he can scan the waste, and the robot will tell the person if it is recyclable. The robot has a container on its body where the user is able to throw the recyclable waste.

Design and the Prototyping Methods

The Body — Model Prototyping

This method was used to prototype the body. Basically, we tried to make this robot look the same with Wall-E. The body was made of wood boards, the eyes was made of white boards and 2 cups, the wheels was made of two rollers for 3D printing filament, and neck was made of a paper towel tube.

The prototype is used to evaluate how people would response to the appearance, if the size of the robot is proper, and if the positions of the “scanning sensor” and the container are intuitive for the users.

Mobility — Electronic Prototyping

In order to make this robotic trash can move, we used 2 motor wheel to move the robot’s body and another three supporting wheels to ensure it stable during the movement. In order to make sure it would not run into the wall, we added a distance sensor. When there is something in front of it, it would move backwards and then turn left.

The robot is also remotely controllable. We used a pair of Xbees to wirelessly connect the Arduino and our laptop. In addition, we also launch an application called CoolTerm and control the robot by typing the commands on the app— “walk forward”, “move backwards”, “turn left”, “turn right”, and “stop”.

This prototype is used to test that, if the robot moves, how people would response. What are the behaviors when the users see the movements. In this way, we can also test the speed and even the motor.

Scanning — Wizard of Oz

We put a LED light on one eye of the robot and implied the eye has the sensors to scan the waste. We put a phone inside of Wall-E’s eyes and use Skype to lively monitor what people is scanning in front of the eyes. In this way, we can remotely control Wall-E what to speak to the user.

The purpose is that we created the robot to be like a true product, and let the users believe it is really working as the way we showed, so that we could evaluate people’s behaviors.

Speaking — Wizard of Oz

The robotic trash can walk around, stop by the users (we see it through Skype and control it remotely) and ask them “Hi! Do you have anything to recycle?” If the persons answers “yes”, it would reply “Great! Please show me the item.” At this time, the person may put the waste in front of Wall-E’s eyes. If the waste is recyclable, the robot would reponse “You are right! Please place it in my bin.” If it is not, it would reponse “would you like to try another item?” After all the process (until the user says “no”), the robot would just say “Thank you! Bye-Bye!”

We put a wireless speaker inside the robot and played the audio on the laptop. As mentioned, we monitored what was happening in front of the robot so that we could decide which audio to play according to the context.

The purpose of the this prototype is the same as “Scanning”, which is evaluating the users behaviors.

Demo — Video Prototype

In order to demonstrate how the robot works, we created a video to show the functions and how it interacts with a user. In the video, we also included usability test process. Please visit our video:https://www.youtube.com/watch?v=q6ciKACiLKs&feature=youtu.be

Usability Test

Because our target user is UW-ers, we went to the Odegaard Library to evaluate people’s responses and interactions with the robot. First, many of them were excited with the new product showing in the library. Over 30 people were excited to try the product and did the usability test with us. Including the final demonstration in the dub seminar, over 60 users helped with the evaluation.

Many users believed the robot could sense their needs and the scanning and speaking was “real”. Some of them was curious about the new technology. Most of users were clear about how to interact with the robot; a couple of the users thought we should make a clear sign for the scanning area even though they immediately figured out that they should scan in front of the eyes because it is intuitive. In general the design was effective; all of the users thought a robotic trash can was fun to interact with — the robot collected a lot of trash when it came back.

Lessons

In this project, we learned how to integrate different prototyping methods comprehensively. In the model prototyping method, we learned how to reasonably choose materials for different body parts; we also experienced how to properly adapt current technology to wizard of oz method; In the electronic prototyping method, we learned the logic of building the robot mechanism and how to choose and use different sensors and the Bluetooth wireless technology. Most importantly, we learned how to plan and finish the project with the limited time and cost budget.

We faced two big challenges about technology: wireless control and test on wheels. The Bluetooth we firstly ordered online did not work properly and we could find a Bluetooth at a local store. Later, we learned how to use Xbee to prototype it. Second, the initial model prototype had two motor wheels and one supporting wheel; however, the robot was not very stable in its movement. Therefore, we added two more supporting wheels in front to ensure it moves smoothly and stably. In addition, the speed of the robot changing directions could be affected by different floors. For example, the robot could turn left by 90 degrees on carpet, but may be 180 degrees on hard floor.

We were especially surprised how effective and powerful the wizard of oz method could be. Many people believed it is a real robot with actual technology implemented on it. When we brought it to the Odegaard library, many people thought robot was real and took pictures for it. We thought the project was successful because people really loved to interact with the robot.

Analysis of the Techniques

There are so many techniques fitting into our personal toolkit: Modeling technique, Arduino, sensors, wireless controls, and even time management technique. We found one of the pros of Arduino was that we could use many shields that provided us flexibility of building the system. The wireless was actually not hard to set up on it and connect it with personal device. The cons we found on the Arduino was the libary. Even though there are many libraries available, when setting up the first speaker, the library was confusing. But, Arduino is so powful that we can use it on many kinds of applications, for example arduino wireless can be used for remotely controllable toys.

What We Learned about the Design Process

Planning ahead is the key of design process. It can save much time for us if we planned and prepared very well. When we are not sure whether the function can be realized as our expected, we have plan B and plan C, which we found really important later. In case of the voice of speaker on Arduino is too low, we prepared a bluetooth speaker. What’s more, we found iterative evaluation very helpful to improve our design.

It has stretched both our technical skills and creativity by learning using new sensors and express my ideas in a physical form. Overall we have been pretty excited about implementing skill sets we have learned this quarter in the studio and integrating them into the final project.