P2 — Elite Four

Posted on March 11, 2013 by Jae Young Lee

Group Name and Number: The Elite Four; Group 24

Names and Contributions: Clay Whetung, Jae Young Lee, Jeff Snyder, Michael Newman

Clay conducted one of the interviews and helped answer many of the questions, especially regarding contextual inquiry.
Jae conducted one of the interviews and helped answer many of the questions, especially regarding contextual inquiry.
Jeff conducted one of the interviews and helped answer many of the questions, especially regarding task analysis and interface design.
Michael drew up the storyboards and sketches and helped answer many of the questions, especially regarding task analysis and interface design.

Problem and Solution Overview

We are addressing the problem of people (especially students) leaving important items (e.g., keys, wallet, phone) behind when they leave their rooms/buildings. This can lead to various further problems, like being locked out, being unable to start one’s car, or being unable to send/receive texts or voice calls. Our proposed solution is a system that uses proximity sensors to detect when a user is attempting to leave a room without their keys/phone/wallet and alerts them (visually or audibly) before it’s too late. Our sensors can also help a user find missing items around the house or even out in the world. This addresses the problem by both preventing and dealing with the aftermath of user forgetfulness.

Description of Users Observed in Contextual Inquiry

Our target user group includes all people who leave important items behind. Specifically, we are focusing on rooms/buildings with self-locking doors, since the consequences of forgetting something are often more severe. A great example of this user group is college students. We also have easy access to these students and can make first-hand observations without much trouble. The first user we observed lives in a quad in Brown Hall. He is a junior in the COS department who describes himself as fairly organized. He spends a large amount of time in his room since it is one of his main places for studying and doing work. The second user we observed was a CBE senior living in a triple in 1903 Hall. He was generally organized, but his room had become quite messy in the past few weeks as his thesis workload increased. He often leaves his room to work, go to the gym, do laundry, and attend various extra-curriculars. He would like to be able to ensure that he is completely prepared when he leaves the room, with minimal effort on his part. The third user we interviewed lives in a quad in Little Hall. She is a senior in the PSY department and described herself as organized, though her room was somewhat messy. She would like a solution to help prevent expensive lock-outs.

Contextual Inquiry Interview Descriptions

We interviewed the three users in the environment that our system would be employed — their dorm rooms. Each of the three was observed leaving their rooms. We logged the habits that they had formed for preparing to leave their room. Afterwards we discussed with each participant their routine in detail and how they thought their current system could be improved. We also asked them what they were willing to sacrifice for an entirely new system and provided them with some of our initial ideas for our project in order to help focus and guide the discussion.

There were several common experiences that each user shared. They all commented on the habits that they used to ensure they were ready to leave their room. This universally involved checking pockets/bags for wallets, phones and keys. However, these systems fail catastrophically when key items are missing from the wallet, or items are mistaken for each other. For example, users would remove their prox from their wallets to do laundry or go to the gym, and when they forgot to place it back into their wallet they were locked out. Along with this, they would lock themselves out if they felt their phone in their bag, but mistook it for their wallet. These were by far the most common lockout cases. They also commonly propped their doors to prevent lockouts, but this could lead to fines by fire safety. Two of our users also expressed interest in a device that would help them find lost items. Such a device would preferably be stationary, such that it does not become lost as well. Interviewees also stated they they would only use a system that was easy to install and did not change the form factor of their necessities very much.

Interestingly, one of the users that we had interviewed had investigated permanent solutions to the door locking problem. They had set up a variety of systems in their room to allow access without their prox, so they could never be locked out. However, this was a massive security issue and they were fined multiple times for their efforts. One user was willing to go to great ends for this system, even briefly considering total room overhauls for an optimal system. However, they did express that such changes may be unfeasible.

Answers to Task Analysis Questions

1. Who is going to use the system?
We’re focusing the system on students, but it’s usable by pretty much anyone who lives in a house/dorm/apartment and owns keys, a phone, a wallet, a purse, or other important items.

2. What tasks do they now perform?
Currently, users are simply forced to remember everything they need to bring when they leave home. When they want to find missing items, they must perform either an inefficient sweep over their entire room(s) or a slightly better search of possibly incorrect “last seen” locations. Finding missing items outside of one’s home is pretty much a hopeless task.

3. What tasks are desired?
Ideally, users wouldn’t have to rack their brains for missing items every time they leave the home. Instead, they will be automatically reminded if they’re about to leave something behind. Searching for missing items, particularly in one’s own home, should be faster and easier than simply looking everywhere.

4. How are the tasks learned?
These tasks are mostly learned by habit; they are performed several times a day every day by users. Most users develop strategies for finding lost items early in life, though they may not confront the problem of remembering important items until they live on their own for the first time.

5. Where are the tasks performed?
A user will generally attempt to determine if he has all necessary items immediately before leaving their abode, usually when close to the exit. Searching for missing items can occur anywhere, both inside the home and out.

6. What’s the relationship between user & data?
We don’t collect personalized information about the data or create a centralized data store; therefore privacy concerns are minimal/nonexistent. The only data we collect is proximity data (e.g., are the user’s keys nearby?), which doesn’t need to be stored.

7. What other tools does the user have?
The user’s memory is their primary tool, both for remembering not to leave things behind and for trying to find missing items. However, the memory is a fickle and unreliable tool. For finding one’s phone in particular, there do exist apps that use GPS or other mechanisms to track the missing phone. However, these tools may not work in all situations — for example, if the phone is powered off. For remembering items and being let back into a room after being locked out, a roommate might be a useful “tool.” Many users use hacked solutions — makeshift door stops, door mechanism modifications, or others — to prevent their doors from ever closing.

8. How do users communicate with each other?
When locked out, Princeton students will call Public Safety with their phone, if they have it. If phoneless, they will generally attempt to borrow a phone from a neighbor or kind stranger. Roommates who are locked out may contact each other via cell-phone or email to borrow each other’s proxes. If users realize they have forgotten other important devices while away from their rooms, they may contact their roommates or significant others and request that they meet up and bring the forgotten devices.

9. How often are the tasks performed?
Leaving a room is done multiple times per day. Finding missing items (hopefully) is done less frequently, depending on the forgetfulness of the user. On average, the users interviewed and our group members get locked out on the order of once every 2-3 months.

10. What are the time constraints on the tasks?
There isn’t really a “time constraint” on not forgetting to take one’s keys/wallet/phone out of the room, although usually this is a process that occurs within just a few seconds. On the other hand, finding lost items (especially for important things like phones) is a task that is best accomplished within a short time frame — such as a half-hour to a couple of hours. If a user ends up locked out, they may need access within a short time frame if they have left necessary items in their abode, such as completed assignments.

11. What happens when things go wrong?
At worst, a user will be locked out of their room (potentially wearing nothing but a towel) with no way to call someone else for help. Less drastic scenarios include simply being locked out, being unable to start a car, being unable to open a door or a bike lock, being unable to make calls or send texts, and/or being unable to purchase things. Some of these scenarios could be seriously problematic; others are merely annoying.

Description of Three Tasks

Task 1: Being reminded to take keys/phones/wallet along when leaving a room.
Our system will have a proximity sensor and be located (by default) near the door. When a user opens the door to leave their room, our system will flash red LEDs and play a warning melody if certain items (e.g., keys, phone) are not detected. If these items are present, the system will light up green (and may also play a happy tune).
Current difficulty: Easy
Proposed difficulty: Trivial

Task 2: Being able to use the proximity detector in one’s house to find missing items.
Though our system will typically be located near the door, users will also be able to carry it around (with some kind of battery pack) in order to detect important items in their home. Since the area being searched is relatively small and confined, it should be relatively easy to find objects simply using proximity sensing. The device will have an “item-detecting” mode that the user can activate, and it will light up green and beep more quickly as it gets nearer to the missing item(s).
Current difficulty: Moderate
Proposed difficulty: Easy

Task 3: Using the proximity sensors to find missing items out in the world.
This task is very similar to the previous task in that it involves detection of missing items; however, finding missing items outside of the home environment is far more difficult due to the increased size of the area being searched. For this, the user will need to have a general sense of where their item(s) might be; unlike in the previous task, a broad sweep over the entire searchable area will not be feasible. Battery life and detection range are also more of a consideration for this task; we may need to use sensing devices with a longer range than (for example) RFID tags.
Current difficulty: Hard
Proposed difficulty: Moderate

Interface Design

Our system consists of several parts: tags of various form factors that users can attach to important items and a door-frame mounted sensor to check for items when leaving the room that can also be detached and used as a handheld proximity sensor to locate lost items. The sensor will have a Li-On battery, rechargeable over USB, so it can be used as a portable device. We believe that for wallets, a credit-card form factor tag would be minimally intrusive, whereas for keys and cell phones, a small fob (about the size of a quarter) would be more appropriate. Each of these tags would contain a battery and an antenna. Our system alerts users when they attempt to leave their room or abode without every item they’ve tagged using our system by playing an alert noise and flashing red LEDs. When a user leaves the room with important items, the system plays a happy noise and flashes green LEDs. To add a new item to the set tracked by our system, the user presses a button on the door sensor and holds the new tag up to it until the sync completes. A user can also remove an item from this set by a similar process. The visual and audible reminder provided by our system will help even users in an altered mental state remember their important items. The system can handle multiple users by maintaining separate device profiles for each inhabitant. It will alert a user leaving if it does not detect a complete set of devices. Users can separately add and remove devices. By removing the sensor from its door frame mount, the system will automatically switch to proximity sensing mode so that it can be used to locate missing objects — both in the user’s room and elsewhere. When no device tags are detected the system will flash red and beep slowly; as the user nears the missing item(s), the system will flash green and beep with increasing frequency. As far as we know, no other system grants this kind of functionality to a user — at best, there exist ways to find lost phones using GPS, but only when the phones are powered on. Our system is far more versatile than that, especially since it implements preventative as well as remedial solutions to the problem of forgetting important items.

Storyboards

Task 1: Being reminded to take keys/phones/wallet along when leaving a room.

A clueless user about to leave his room without his keys.

Our system notifies the user that he has forgotten his keys.

Task 2: Being able to use the proximity detector in one’s house to find missing items.

A sad user unable to find his keys.

Our system can be used as a proximity sensor to help locate the missing keys.

Task 3: Using the proximity sensors to find missing items out in the world.

A helpless user who has just realized that he lost his phone at the beach.

Our system can be used as a proximity detector anywhere to find the missing phone.

Design Sketches

A sketch of the proximity sensing device.

How different items would be tagged by the device.

The door-mounted user interface.

The portable user interface.

Assignment 2 — Jae Young Lee

Posted on March 2, 2013 by Jae Young Lee

Observations:

I observed a bunch of random people and purposely did not get their names or tell them I was observing them.

2:20 to 2:30, McCosh 50, before ECO 100 lecture on Thursday, 2/21 (about 300-400 students)

The professor, Harvey Rosen, was there before 2:20, and he spent his 10+ minutes drawing graphs and figures that he would need for lecture on the blackboard, as well as setting up his PowerPoint slides.
A large number of students (I would approximate about 200) were doing nothing other than chatting with friends or sitting in silence.
Lots of students would walk into the giant lecture hall and spend some time just looking for their friends or walking around to find someone they knew. This caused a decent amount of traffic in such a large class.
One student was coding on his laptop during the 10 minutes.
One student mentioned: “I wish I could be doing my math homework right now, but my textbook is back in my room.”
Lots of students pulled out their phone to catch up on texts, emails, and Facebook.
A surprisingly small amount of people (I could only count about 20 on the main floor of McCosh 50) pulled out their laptops. I would guess that laptops are not allowed during lecture, or at least not ideal for taking notes because of the large quantities of graphs in that class.
I left immediately when the lecture started.

1:22 to 1:30, Woolworth 105, before MUS 103 precept on Tuesday, 2/26 (about 15 students)

The preceptor had a class in the same room beforehand, so there were some students who stayed behind to ask questions, which took up about 5 minutes of the preceptor’s time.
The preceptor spent his remaining time writing a few things on the blackboard that he would need later on in class.
One student walked in, thinking he was in the right precept. He looked around and quickly realized he was an hour early for his actual precept, and he scurried out, looking slightly embarrassed.
The preceptor started taking attendance at 1:31. He does it the old-fashioned style by just calling out names (since there are only about 15 people). However, there were 4 students that walked in while he was taking attendance, so whenever one of them walked in, he had to check if he had marked them as absent and then fix it if he had. It seemed highly inefficient.
A good amount of students pulled out their phone during the waiting time.

2:54 to 3:00 in CS 104, before HCI lecture on Tuesday, 2/26 (about 60 students)

Dr. Vertesi (guest lecturer) spent pretty much all of those 6 minutes fiddling with the projector, the lights, and other technology in the room. A student helped her and seemed to get it working, and the lecture didn’t start until 3:03 because of technical difficulties.
A good amount of students were eating/drinking things. I observed a lot of coffee/tea, a banana, a churro, and a granola bar.
The auditorium is terrible for seating: people seem to naturally like sitting towards the edges of each row, which makes it really difficult for people to get to open seats because of how narrow the rows are. I observed about a half-dozen students even climbing over seats just so they wouldn’t have to squeeze through a bunch of people. This problem is made worse by the fact that students want to sit next to their friends.

Brainstorming:

I collaborated with Gabriel Chen.

Hand warmer mats on desks, activated when they sense force.
Attendance (done quickly and automatically, perhaps by proxing in) at entrance of lecture hall/precepts for classes that take attendance.
Attendance (done quickly and automatically, perhaps by proxing out) at exit of lecture hall/precepts to discourage people from leaving early.
An app that automates taking attendance by using GPS: you “check-in” at the beginning of class and “check-out” at the end of class so that the professor knows that you were both on-time and stayed for the entire lecture.
Sensor on every seat that connects to an app that tells you where your friends are and whether the seats are occupied.
An app that has a general map of the classroom that you can use to “reserve” seats for friends; this is useful because it can be awkward telling other students that seats are “taken” when they aren’t actually “taken”.
An app that shows you the outlet locations in lecture halls and whether they’re currently being used or not.
An app that lets bikers press the crosswalk button automatically (so that they don’t have to get off the bike) by using proximity detection techniques.
An app that electrocutes you (really gently…) when you start dozing off in lecture or even before the lecture starts: detects “nodding off” by using an accelerometer on your head.
A live chat with OIT that is accessible on all computers built into lecture halls (for example, in CS 104, both Professor Fiebrink and Dr. Vertesi have had issues with the technology in that room even though they are obviously very capable with technology).
An app that releases Febreze or some other subtle odor cancelling substance when it detects that you are farting or burping in a lecture hall.
A device or system that lets professors see what they’re writing on the board from the back of the room perspective, so they can determine how large to write on a blackboard.
A portable device that can project things clearly onto a blackboard: some professors spend the 10 minutes before class writing material or drawing graphs that they will use later on during the lecture, and they need to use the blackboard because they have lecture slides to show on the projector screen.
A device built into the top of a blackboard that releases a torrent of water so that professors can erase the board from a previous class with a gesture or touch of a button; there is also a drain at the bottom, and a dryer for the board.
A system that allows for delivery from restaurants directly to your lecture hall so that they’ll be there with your food by the time you get there.
A device that automatically scrolls your phone if your eyes are at the top or bottom of the screen; students are often carrying lots of things and don’t have a free hand to actually use their phone.
Phone keyboards that sense where your hand is positioned so that when you want to type with one hand, the keyboard is condensed onto one side of the screen; this is another problem for students who are carrying a lot of things to class.

Ideas chosen for prototyping:

An app that automates taking attendance by using GPS: you “check-in” at the beginning of class and “check-out” at the end of class so that the professor knows that you were both on-time and stayed for the entire lecture. For classes that take attendance, I have never seen a good system for doing so. Some classes pass around a sheet, which often takes a really long time and inevitably skips over some people because students decide to pass the sheet in whatever direction is most convenient for them. Some students also come just to sign-in, and then leave. This app would automate the attendance process and also help discourage students from leaving lecture.

A typical iPhone screen. Attendance Made Easy is the name of my app.
Users need to sign in with their Princeton NetID and password to verify their identity. They can stay signed in to make it easier to use day-to-day.
The home screen for the app itself. There is easy-access to turn GPS services on/off since GPS is required for the app to work properly. There are three main buttons for signing in, signing out, and leaving a comment for the professor.
The sign-in screen. Based on GPS, it tells you what building you’re located in so that you know you’re signing into the right class. There’s an option to recalibrate if the GPS reading is wrong.

The sign-out screen. Based on GPS, it tells you what building you’re located in so that you know you’re signing out of the right class. There’s an option to recalibrate if the GPS reading is wrong.

The comment screen. It tells you what class you’re in and the professor that will be receiving your comment. Basically, I envisioned using this option for any extenuating circumstances if a student was late or needed to leave early but had a legitimate excuse.

The waiting screen if you click recalibrate. Basically the GPS would just refresh your location.

The screen you see after successfully signing in. There’s just a single button to exit the app.

The screen you see after successfully signing out. There’s just a single button to exit the app.
Phone keyboards that sense where your hand is positioned so that when you want to type with one hand, the keyboard is condensed onto one side of the screen; this is another problem for students who are carrying a lot of things to class.This seems like a feature on phones that should already be implemented. As phones have been getting bigger and bigger, it’s become increasingly harder to type with just one hand (especially for people like myself with small hands).
This is what the normal phone keyboard looks like when it senses two hands using the phone (the default keyboard, no different from a phone in its normal state).
When it senses that you take your left hand off, the keyboard becomes condensed on the right side so that all keys are easy to reach with just your right hand. For users who have big hands and have no need for this feature, they can press a button to revert back to the default keyboard.

When it senses that you take your right hand off, the keyboard becomes condensed on the left side so that all keys are easy to reach with just your left hand. For users who have big hands and have no need for this feature, they can press a button to revert back to the default keyboard.

User Testing:

Osman Khwaja, a junior in the COS department, tested the attendance app.

He said there was too much text. For the sign-in and sign-out screens, he didn’t want to bother reading all of that. Just have a couple words that list the location, time, etc.
He never used the comment button. He didn’t really know what it was for, and wasn’t really sure it was necessary even when I explained what I envisioned it being used for (a student with an excuse for being late or leaving early).
He said that the app would be useful overall and make things easier.

Ross Smith, a sophomore in the ECO department, tested the attendance app.

He said he would never use the comment button because there are already easy ways to interact and communicate with professors. Afterwards, I told him I envisioned it being used for students who had an excuse for being late or leaving early, and he said that made more sense. He suggested just renaming it to an “Excuse” button.
He said it’s obviously not a viable system unless everybody has a smartphone.
He didn’t really understand how it worked when he was first going through the home screen and sign-in screen. When I explained that it knew what classroom you were in through GPS, he immediately understood the entire app a lot better.
Overall, he thought it would be a very streamlined way for professors to take attendance and make sure that students stay for the entire class.

Stephen Wang, a junior in the ORF department, tested the attendance app.
GosmsPhoto1362197752885

He actually knew what the comment button was for. He said he would use it to explain to the professor why he was late. He said it would obviously not help for telling the professor why you’re absent since you can only use it if you’re in the right classroom.
He also mentioned that it requires all students to have a smartphone.
He said that it would motivate him to go to class more often, which is a great side effect.
He said it was easy-to-use and a really good idea.

Insights:

Each button needs to have a clear function and a simple yet descriptive name. I created the comment button thinking students would want to leave comments to professors if they had any relevant issues with their attendance that day. However, when users saw the word “comment,” many of them just assumed it meant general feedback for the professor about the course, and they were confused about the functionality or just didn’t bother using it. I should rename the button to say, “Leave the professor an excuse.”
People are lazy. They don’t want to read a lot of text. The app should be as streamlined as possible since it’s something that students would potentially use multiple times a day.
It’s unrealistic to implement this for actual attendance since it requires all students to have a smartphone. I kind of realized this from the beginning, but still wanted to prototype it and get feedback.
There’s potential. My testers all said it would be useful, and that there’s really no good way for professors to take attendance right now without being inefficient or annoying in some way.

Human-Computer Interface Technology

Princeton University

Author Archives: Jae Young Lee

P2 — Elite Four

Assignment 2 — Jae Young Lee