# Team Plan H : Low-Fi Prototype
## Problem Statement

Exercise environments are reduced due to Covid-19 pandemic. Since the number of available exercise routines decreased and those left are too strict, one becomes less motivated with the exercise. It also became harder to find suitable and enjoyable routines due to the limitation of environment and tools.

People usually share their exercising experiences through SNS, by posting pictures and short text because it makes them motivated. However, this does not exactly show what kind of exercises they have undergone in detail.

It is hard to get help from experts nowadays. Even if people find their own routine, it is hard to develop it further without advices from experts. There is almost no way for non-experts to share one's unique routine and develop it by combining others' expericence.

## Task

1. **Easily** write and share one's own routine.
- Users can easily share their exercising experience or plan in the simplest format.
2. Explore **diverse** routines of user-created tags.
- Users can search suitable routines from diverse routine library by tags that are not formulated.
3. Leave **comments** and specific **suggestions** for update in user review.
- Users can communicate interactively by reviewing.
4. **Update** routine based on suggestions supported by the majority.
- Non-expert routine can be developed by combining experiences of lots of people.

## Prototype
#### [Link to prototype](https://xd.adobe.com/view/0ec1ff81-c633-44de-bb57-f4050a90ff98-5b92/?fullscreen&hints=off)
#### Prototyping tool : Adobe XD
#### Why?
We used 'Adobe XD' after comparing with other suggested tools 'Figma', 'Marvel', 'InVision', or 'proto.io'. Some of our members already had some experience with Adobe XD. We noticed that Adobe XD supports multi-person work on the cloud very well. Adobe XD also provided various functions related to the design itself.

#### What worked well with the tool?
For designing, functionality for modifying same entities at once was supported and it was easy to aggregate individual designs.
Designing features was also satisfying, especially the interactions to finish prototype was very well organized in the tool. Diverse interaction types were supported and they were easily expressed with recognizable formats of arrow.
#### What didn't worked well?
Version controling between members was one of the problem. Also there were problems with copying and modifying same assets.
Drawing functionality was not satisfying. One who are not experienced in vector-image drawing had difficulty drawing smooth curve.
Dragging function works well in the Adobe XD file, but it does not work well when it is exported into link.
### Design choices
1. **Easily** write and share one's own routine.
* We did not implement actual typing experience due to the limitations of Low-Fi prototyping tool.
* We did not implement presets or tools to help user write the routine more easily, which is not the key feature of our project.
* We did not provide a detailed tutorial or guideline for routine posting, which would be neccessary in the real product, since we provide the instructions seperately.
* We did not implement comfort features regarding multimedia such as inserting video by url, which is not meaningful in the step of Low-Fi prototype.
2. Explore **diverse** routines of user-created tags.

* We did not implement increasing or decreasing the time by clicking triangles because it was trivial at this moment.
* We did not implement autocomplete when user input some texts for applying tag, becaues of the limits of low-fi prototype. We just showed some samples that users can click on. It would be implemented in high-fi prototype with more details.
* Users can also check their exercise levels by answering several question. We did not implement algorithm computing appropriate level with given answers because there was not enough user data to support this result.
* We did not implement manual algorithm for several results and just showed one example, because it was hard to support every case and prepare examples. Our purpose was to show simple searching methods with diverse tags.

3. Leave comments and specific suggestions for update in user review.
* All the reviews and profiles were fake and most of them were copied and pasted. We thought just showing how it will actually look on the screen was enough.
* Starting a routine menu is not implemented. Although we thought that this is a common functionality that every training assistance service should provide, it was far from our unique tasks that should be shown in advance in Low-Fi. This menu will be implemented with detailed description of steps later on.
* Adjusting scores in pentagonal graph function was not implemented because user did not actually experience exercise. When implemented, adjusted scores would be averaged and shown after the review submission.
* There should be a functionality for sorting reviews. However it didn't seem to be a unique feature to implement in Low-Fi, so we left it for later on implmentations and only showed filtering by comments or suggestions.
* When filtering only comments or suggestions, detailed graph with rating scores become unavailable. This was due to an assumption that those who use filtering would focus on the content of the filtered results, rather than pre-existing rating scores. We plan to implement functionality for active reviews later on.
* In the Low-Fi you should submit both comment and suggestions and they cannot be separated. This is due to the restrictions of Low-Fi, and users would be able to write only what they want in higher fidelity prototypes.
* Other functionalities like up/down voting, reporting, edit/delete and some back buttons are shown in the prototype but doesn't function. Implementing those individual functionalities were less important and might confuse the participants. Those dummy buttons would function as they were intended in later works.

4. Update routine based on suggestions supported by the majority.
* Dynamic interaction with user input was difficult to implement in Adobe XD so updating feature followed the steps predetermined by developers.

5. General
* We did not implement link for every user profile image connected to profile page since it was too complicated to link every button.

### Representative screenshots
▲ The steps for writing a routine is easy and straightforward.

▲ User can search with time, level, tags and get recommendations for levels after answering some questions

▲ There are various ways to respond to the given routine.


▲ The comments and suggestions are shown on the side when user is updating the routine.

### Instructions
1. Set up your profile first.
a) You don't have a profile yet. Let's set up your user profile.
b) By clicking slots unfilled, our prototype would respond to your actions.
2. Now let's write a new routine.
a) Start writing a new routine.
b) Set the name and tags of your routine.
c) Add a new action.
d) Add information regarding your action.
e) Post your routine!
3. Try to search for a routine.
a) Start searching
b) Set the time span you want to exercise
c) Check your level to know your exercise skills
d) Fill up your levels as recommended.
e) Add "stair" and "dog" tags.
f) Now see the result.
4. Browse and write comments, update suggestions.
a) Many routines are shown! See more detail of a routine.
b) There are some comments in routine! Browse comments in detail.
c) Try dragging down the comments
d) See only comments or only suggestion to browse what you want
e) Write your review! Submit your suggestion and comment.
f) Now see your comment.
g) Return to timeline.
5. Update your routine.
a) See your routine and go to update.
b) Now you can see on the right some suggestions written by others. Edit your routine accepting the suggestion.
c) Now upload your modification!

## Observations
### Participant
#### Participant 1
- Mother of Jisu Yim. (52 yo)
- She exercise regularly doing yoga and home training seeing youtube video. She attended to yoga academy but due to COVID-19, she sometimes cannot go to there and was forced to do exercise in home for long time.

#### Participant 2
- A male KAIST student majored in computer science school. (22 yo)
- Not much experienced in exercise, but trying to start with several workout manager application. However, determination easily fades out.

#### Participant 3
- A male web designer (32 yo)
- Doing exercise regularly, 90 min per day, 3 days a week.

#### Participant 4
- Brother of Taehyung Kim.
- Majored Industrial Engineering in Cambridge, England. (24 yo)
- Currently working in Medical start-up.
- Interested in online platforms.

#### Participant 5
- A male KAIST computer science major student. (23 yo)
- Recognized importance of exercise and currently finding platforms to start with.

#### Participant 6
- A male KAIST computer science major undergraduate student. (26 yo)
- Workout regularly and have interests in professional trainer accompanied trainig.

### Usability Problem

1. Easily write and share one's own routine.

* **Medium** It would be hard for beginners to check their posture with only given posting. Interactive feedback on wrong postures would be neccessary. (P5)
* **Medium** Adding every instruction one-by-one seems uncomfortable. (P6)
* **Low** It is not favorable to read because most of the contents are only written in text. (P2, P3)

2. Explore diverse routines of user-created tags.
* **High** The concept of tags are somewhat vague, how they are created and searched should be clear (P2, P3, P4, P5)
* **Medium** Setting time and difficulty range should be more intuitive (P4, P6)
* **Low** Search button should be seperated from toggle button. Currently search icon performs both functions. (P1, P2, P4)

3. Leave comments and specific suggestions for update in user review.
* **High** There should be an organized way to normalize difficulty of exercise. (P3)
* **High** How could the quality of review guranteed? (P6)
* **Low** Sorting and filtering function should be implemented. (P1, P2)
* **Low** Some rating criteria seems to coincide with the purpose of overall rating. (P6)

4. Update routine based on suggestions supported by the majority.

* **Medium** Reviews and ratings should be modified corresponding to the updates. How are these considered? (P6)
* **Low** Hard to know that profile image is linked to see my routine. (P1, P2, P3, P4, P6)
* **Low** Why does "UPDATE" button suddenly appear up without any context? (P3)

5. In general
* **High** Service doens't seems to motivate users. It would be motivating if system asks daily routine progresses or having competing other users. (P5)
* **Medium** Words should be unified, such as (profile, timeline, my page), (review, comment). (P4, P6)

### Plan

1. Easily write and share one's own routine.
- Additional feedback and cautions could be provided for the beginners and the list of cautions could be continuously updated by other users.
- There should be some presets for writing commonly used instructions.
- Adding image presets for widely-known postures would help.

2. Explore diverse routines of user-created tags.
- Tags should be organized so that it is clear whether user created one or it was formatted.
- Time and difficulty tags required so we can differ colors to emphasize.
- Time tag: We will add range of tags so that the users can find routines that spends 25~30 minutes. Also for users newly writing the routine, those time tags will be supported in range (e.g. 5~8 min).
- Level tag: Verifying level of routine is subjective. Simply thinking, our intention to add questionaire of "How much time can you do plank" was to know the general ability of the user. We will combine information that indicates exercising ability with routines they usually choose, along with the information regarding how they evaluate the level of routines. Also, asking for advices to some experts will be helpful. We should find the most reasonable way of utilizing social-computing method.
- Specific tags (e.g. stair, yoga): For more various results, we can show results that include certain tags. For example, if user add 'yoga' tag, result will include routines that has tag 'easy yoga'. Also, we will support autocomplete function that shows related tags of user input, and some recommendation that user will want to add. If user does not add any tags, we will recommend demanding tags.

3. Leave comments and specific suggestions for update in user review.
- Guarantee the quality of reviews : Rating and review should be also modified after update. For other restaurant reviewing services, they have category for showing reviews from few weeks ago. Thus users can keep up with recent responses. If there are constant amount of reviewers, quality can be assured. We could assure the consent of reviews by filtering old reviews after update of the routine, possibly noted with "suggestion accepted". Also rewarding helpful reviewers can motivate other users to add suggestions actively and maintain the quality of reviews.
* Add submit button and make user to review with only suggestions or comments.
* User's reviews should be visible from the beginning.
* Sorting recent, popular reviews.
* Adjusting scores for the routine evaluation / fixing the criteria.
* Add up/down vote, reporting, editing and deleting, like button functionality.
* Expanding pop up or overlaying detailed review page triggered by clicking simplified button should be added.
* Users should be logged in to leave a review and there will be some constraints to prevent trolling. They should have at least looked through the routine pages. Verified users will be provided with prestiage to review by just looking at routine outlines. But only browsing the routines could be done easily without account.

4. Update routine based on suggestions supported by the majority.
* To enable an user to be informed diversly, both the recent and the entire rating will be displayed in one screen. For a writer, this allows to directly see the effect of update when suggestions were followed so that he can decide to keep or revert the change.
* Functions omitted due to the restrictions of Lo-Fi will be implemented. User will be able to interact with the application more dynamically.
* Difficulty of routines should also be modifiable.

5. In general
* Actual menu for users to follow. It can be automatically translated from routine post formats into followable detailed stepped format. Detailed stepped format will support novice users more carefully. Progresses could be recorded. Additional features helpful for exercising such as background time control and queing(counting numbers) could be added.
* Unify the terminology used. Like "Profile, timeline, ...", "comment, suggestion, reviews", "submit, write".
* Add user account and profile pages. All user profile images should link to that page. User profile page should include list of followers and followed / routines posted / routine history / liked routines.
* Add auto-recommendation for routines. It can be visible at the timeline, during searching process.
* We initially thought that trying other popular routines, posting routines and interactively getting feedbacks can provide enough motivations. However, one of the participant suggested that it might lack those features. Adding competitive features or other motivating aspects like reward system, reputation system or verified user hierarchy might be a solution.