apq3uw

[apq3uw]

Evaluation of Digital Mock-ups

Let's start with some summarized statistics for the Leikert survey questions. In this post, "V1" or "Version 1" refers to the first set of myBalsamiq mockups; "V2" or "Version 2" refers to the second set.

Figure 1: Summarized Statistics

All of these questions used possible values 1, 2, 3, 4, and 5, with 5 "happiest" or "best". Version 2 was preferred by users across the board. This is good, since that version incorporated a number of improvements from user feedback as described in the "Digital Mockup Design Revisions" post on this blog.

The standard deviations are similar but, on average, slightly less for Version 2. I don't think the differences are significant.

Now, let's take a look at the responses to each Leikert question:

Question 1: Please rate the ease of creating a pill-taking schedule

Figure 2.1: Responses to Question 1 for Version 1

Figure 2.2: Responses to Question 1 for Version 2

This was the question with the highest difference in average response score between the two versions. This makes sense as it is where most of the improvements were made. Major improvements included bringing day-of-week selection more into the main flow and adding a schedule-by-day screen. Figure 2.1 shows that one V1 respondent had a negative impression of the schedule module; that did not occur for V2.

One comment on the V2 open-ended questions showed that those changes bore fruit:

- I like the week view and the ability to touch a day of the week

Another comment showed where there is still some room for improvement.

- Easy: add and remove pills. Difficult: the day buttons are very small.

The small day-of-week buttons (used to select which day(s) of week a pill time is applicable) is something I already wasn't crazy about. If the design was taken further, I would look into using a pop-out window for day-of-week selection, which would allow for larger buttons.

Question 2: Please rate your confidence level in the reminder process leading to taking the right pills

Figure 3.1: Responses to Question 1 for Version 1 

Figure 3.2: Responses to Question 1 for Version 2

There was little difference between the versions on this question. Some improvements to the process were made, particularly with not letting the user proceed until they had checked all of the boxes on each screen, but there was not much room to grow.

Question 3: Please rate your overall happiness with the reminder process.

Figure 4.1: Responses to Question 1 for Version 1

Figure 4.2: Responses to Question 1 for Version 2

This question also had a substantially higher rating in V2.  

An open-ended question answer on the V2 survey suggests a reason for the high rating:

- "I like that the buttons are large and are in different colors so that they are easily distinguishable."

But, these aspects were also there in V1. It is possible that the "Reminder" results were influenced by the "Schedule" results since the Schedule module was presented first in both the mockup links and questions. Of course, there may also have just been some randomness due to the small sample size. 3.75 is within one standard deviation of both mean values.

Here are the Desirability Test results.

Figure 5.1: Desirability Test responses for Version 1 

Figure 5.2: Desirability Test responses for Version 2

These results further illustrate the general preference for Version 2. For Version 2, no words that I consider "negative" were selected while several such words were selected for Version 1.

I was happy to note that "Accessible" was selected by all four Version 2 respondents as accessibility was a key goal of my design. It still tied for the most frequently (50%) selected word for Version 1 as well.

Overall, the Schedule Module Leikert and open-ended question answers pretty clearly align with the narrative of "the improvements from V1 to V2 made a difference". The Reminder Module results also showed improvement from V1 to V2, but the reason why is not as clear. The desirability test results further illustrated the general idea that V2 is better, with V1 having a mixture of positive and negative words selected and V2's word selections only being positive.

[apq3uw]

A Business Model for PillMaster

The value proposition for PillMaster is that it can take the stress out of what is commonly a source of ongoing low-level stress for many different kinds of people.  While pill management is often associated with the elderly - and they are certainly a key user group for the product - anyone who takes any kind of pills regularly could benefit from the product.

PillMaster can address several different sources of stress / uncertainty that are caused by pill management.  

It addresses the “I forgot to take my pills” problem by creating audio / visual reminders (similar to an alarm clock) at the appropriate times and continuing to send alarm signals until receiving user confirmation that pills have actually been taken.

It addresses the problem of taking incorrect pills by waking the user through exactly which pills to take at a given time and requiring that the user confirm the name and quantity of each pill taken before ending the session

The PillMaster physical pill container (not emphasized in the recent round of digital mockups, but part of early storyboarding / user interviews) would make it easy to see which pills have been taken and which pills should be taken next.

The key activities for PillMaster are tied closely to the sources of stress described above in the value proposition.  It provides reminders (repeatedly if necessary) every time pills need to be taken.  It provides a detailed list of pills and quantities to take with each reminder.  Lastly, the circular design of the pill container makes it is easy to see which pill sessions have been completed, which one is next, and which ones are to come.  (This information is available in the mobile app as well, though an additional screen would probably make it easier to get that specific information.)

Success metrics for PillMaster are centered on user adoption.  We need to know not only how many people have the app installed, but also how many have actually set up schedules and are using the reminders to assist in their pill-taking.

Additionally, to guide future product distribution, we need to know how users are getting the product.  Much of my revenue stream analysis focused on getting the product to users via people / companies that are interested in possible users taking pills properly and, in general, being healthy.  Over the past few years, health insurers and employers have started to pay insureds / employees to learn more about their health and adopt healthy habits.  The idea is that people in good health have lower long-term health care costs.  This concept could be extended to pill-taking - people who take pills regularly as planned are more likely to avoid costly ER visits and/or have their health deteriorate.  Family members of people who take pills have a “love your family” interest as well as a secondary financial interest as well.

Health insurers, medical providers, family members, and government agencies are all called out as possible “indirect” distribution channels to get PillMaster to users.  That’s a lot of possible channels and we would really benefit from knowing which of them are working effectively at getting users the product and which are not.  We could then focus our ongoing distribution / advertising efforts on the channels that work well.

Completing the Lean Canvas got me thinking about these distribution channels and possibilities for integrating PillMaster into existing “health management” systems that are already in the marketplace.  Future iterations of the design may include possible “hooks” to include branding from other distribution channels or ways in which people from those channels can be included in the design.  For example, the app could provide contact info of who to call with questions or let you configure family members / medical provider notification when pills are not taken as scheduled.

[apq3uw]

Digital Mockup Design Revisions

Links to revised mockup

All screens: https://apq3uw.mybalsamiq.com/projects/pillmasterv2/grid

Home page: https://apq3uw.mybalsamiq.com/projects/pillmasterv2/HomePage

Summary of critique / TA feedback:

* The use of “Remove” as a label of multiple buttons across screens is ambiguous because different things are being removed.

* There are multiple paths to removing a pill-taking slot, which results in an unclear user experience and takes up too much space.

* Pill quantity is not labeled; the unlabeled quantity is hard to interpret.

* The only Schedule view offered is totally pill-based.  There is no time-based schedule view available and one would be useful.

* The use of different colors for the same thing (i.e. pill names) is distracting and suggests that there are differences between items even though no differences exist.

* It would be nice to be able to enter multiple pills into a single time slot / reminder.

* The “confirm these pills are taken” process is clunky / slow.  It would be nice if the system defaulted all of the checkboxes to “selected” so that the user didn’t have to check them all.

* Buttons would look more like buttons (rather than labels) with some shadow.

* “Done” buttons should only be enabled when required tasks on the screen are completed.

* “Cancel” and “Done” buttons should be better grouped (i.e. on the same line).

* Use of “New Pill” as a screen label is confusing since the same screen could be used to edit information about a pill.

* The “pill type” data entry is confusing because it’s not clear what “pill type” means or how it will be used.

* Some buttons (such as Add Time and Done) are grouped together even though they do very different / unrelated things.

* Associating pill-taking slots to days of the week should be easier.  It is not something that should be “hidden away”.

* Page titles do not stand out from content.

* The list of pills on the “take these pills” screen should include checkboxes so that the user can track taking each pill.

Design updates

I focused on the following areas:

* Add a time-based schedule view.  That seems like a natural user need.

* Get “day of week” into the “main flow” for setting up pill-taking sessions.

* Fix coloring and grouping of buttons so that it better aligns with how the actions taken by each button are related.

* Make page titles stand out.

* Only enable Done buttons when appropriate actions are completed.

Detailed list of revisions:

Home * "I took Pills" changed to "Record pills taken" to be more clear * Bolded title to make it stand out * Added logo

ScheduleHome * Made backgrounds of pill name labels same color * Added splitter to ungroup Add Pill (changed to + Pill to create space) and Done * Bolded title to make it stand out

* Added link to new Schedule By Time page (see below)

Edit Pill ("Aspirin") * Change label to Edit Aspirin to make it clear that we’re on an Edit screen. * Bolded title to make it stand out * Added lines between buttons so unrelated buttons don't seem as grouped * Changed "Remove" label to "Remove pill" to better differentiate from the X buttons that remove that time slot only * Added days of week listing to each time slot * Removed Day of Week button as it was no longer needed.  Added Cancel button, which had been missing. * Added Red X to remove a pill, replacing the "Remove Pill" button. * Replaced the "Remove Pill" button with a Pill Info button to return to / edit the Pill Info screen information for this pill.

Pill Time Edit (the one with quantities) * Removed new vs. now time comparison as it was confusing, low-value, and I needed the space for day of week selection * Changed remaining time label to 'Time'. * Added separators to ungroup dissimilar buttons * Added ability to set days of week on this screen. ** Makes it easier to set days of week (less hidden) ** Allows each time slot to have its own days of the week. - Added pill quantity label - Separated day of week buttons to make it easier to press the right one

New Pill (now called Pill Info) * Moved Cancel and Done to be on same line * Renamed screen to Pill Info so that it can be used as both a "create new" screen and an "edit existing" screen * Replaced "pill type" (wasn't being used for anything) with Additional Info entry. * Removed "+ Time" button.  Time can be added via the Edit Pill screen

Call From Pillmaster * Bolded title.  Changed sizes of other labels to emphasize title further. * Removed partial list of pills - there wasn't enough room for a full list and the partial list was misleading / confusing.

Start Pill Session * Added checkbox for each pill to be taken.  This helps ensure that all pills are actually taken. * Disabled 'I Took Them' button on base screen (without all checkboxes checked) * Created a duplicate screen with all checkboxes checked and the 'I Took Them' button enabled.  This shows that the user will not be allowed to proceed unless they have taken each pill. * Changed color of time label to make it more visible

Confirm Pills Taken * Disabled 'Done' button * Created duplicate screen where 'Done' is enabled and all checkboxes are checked.  You can reach the new screen by checking the unchecked box on the original screen. * Changed color of time label to make it more visible

"All Done" page * Change label from All Done to All Set - less childish * Changed color of time label to make it more visible * Added logo

“Schedule (by time) page

* Added screen for viewing schedule by time.

** Uses accordion menu to select which day of week to view.

[apq3uw]

Digital Mock-up

Links to mockup: 

All screens: https://apq3uw.mybalsamiq.com/projects/pillmaster/grid

Home page: https://apq3uw.mybalsamiq.com/projects/pillmaster/HomePage 

Starting point of “a reminder has gone off to take pills” sequence (also accessible via ‘I Took Pills’ button on Home screen): https://apq3uw.mybalsamiq.com/projects/pillmaster/CallFromPillMaster .

Most buttons with colored background are links.  The screens do not dynamically update (e.g. adding a new pill does not actually add a new pill to the list of pills) but the links let you navigate through all of the screens.

To create the digital mock-up, I relied heavily on my revised UI flow diagram from the Design Revisions assignment.  In most cases, each digital screen was made up largely of the text and controls from a screen in the flow diagram.  I did make a lot of changes to the location of controls, influenced by a desire for consistency, the design principles discussed in the last lecture, and a few “I have a better idea now that I’m looking at this again” items.  Also, two screens that felt awkward in the UI Flow diagram were consolidated into a single screen.

I focused on screen-to-screen consistency, clarity of wording, using large text and bright colors, and creating flow.  My user review indicating varying levels of success with these areas of emphasis.

Examples of good visual design:

- Similarity.  "Good" actions that add data to the app (e.g. set up schedules, complete pill-taking sessions) use green buttons.  "Bad" actions (cancelling an action, removing data) use red buttons.  Actions that fit into neither category use an in-between color.  Among green buttons, the most natural "go forward" / "Done" button is in the sharpest green, most similar to a green traffic light.

- Similarity.  Buttons are placed in similar locations from page to page.  A blue help button is always at the bottom.  On the schedule pages, a green Done button is in the lower right.  On the "pill sesssion" pages, a green "go forward" button is the top button available.  This also helps to achieve contrast.

- Proximity.  Buttons that constitute sets of choices are together, in between vertical barriers.  Thicker vertical barriers are used to separate more dissimilar content; thinner barriers are used to separate mildly dissimilar content.

- Continuation.  Sets of similar items are presented in a common horizontal or vertical format without distracting deviations in the other direction.  Examples include lists of pills names and days of the week.

- Flow.  There is consistent flow from top to bottom and (less so) from left to right.  As mentioned above, each page is anchored with a Help button.  This is particularly important to a product with a likely elderly user base.

- Type size is kept large (usually quite large) to maximize readability for a wide range of users.  Fonts are simple and readable.

- There is almost always some (though sometimes not much) separation between controls to promote readability and keep users from clicking the wrong thing.  An exception is the set of Day of Week options, which was called out by my user as a challenge.

User feedback

Home page

+ Simple.

- Commands don’t seem comparable - “I took pills” and “Schedule” don’t seem like mutually exclusive options.

Schedule page

+ Colors.  Having different colors by pill is a useful visual aid.

- Unclear whether “Schedule” means we are here to take pills or view / create a schedule.

Pill editor page

+ Big red X to remove a time slot is a good visual - very clear.

- “Remove” is ambiguous - what exactly is being removed?  Particularly confusing since there are other buttons that remove something on the page.

Days of week

+ Days of week look like a pill container.

+ Red is a good color for an “off day”.

+ The “all” button will be handy for a lot of use cases.

- Day of week buttons are too small.

Timeslot edit

+ Time picker and AM / PM buttons look very usable.

- “Remove” is again ambiguous (and possibly redundant to the X’s on the Pill Editor page) - it’s not necessarily like the other “Removes” on other screens.

- The quantity menu is ambiguous - it’s to pick the number of pills to take in this session, but that’s not clear.

Add Pill

+ Very clean - good use of whitespace - visual appeal.

- Pill type seems unnecessary and isn’t shown / used anywhere else.  That real estate could be better used by adding something like a “notes” section (e.g. “take with food”).

Reminder

+ “Time to take pills” is engaging verbiage.

- Bell icon would be nice to make it clear that this is an “alarm”.

- Pill list is potentially misleading.  Not enough room to show all pills.

Take These Pills

+ Still can “snooze” if not ready to take pills.

+ Big green button user-friendly.

- Time label is not dark enough - low visibility.

? Might be better to have separate interactions for this pill.  This would arguably lead to better pill-taking results but I thought it would be too chatty.

Confirmation

+ It’s clear what to do.

+ Checkboxes are easy to use.

- Should use present tense verbs.

All Done

+ The “reward” picture is appealing.

+ Showing the date / time of the next pill session is useful.

- “All Done” is a bit childish - maybe “All Set” instead?

[apq3uw]

Design Revisions

I revised my design by doing a new UI Flow Diagram.  The new diagram contains screen revisions and a few new screens.  Compared to my original diagram (also included in the submitted PDF file), the new file includes the following changes.

1) Added help link to Pill Type selection.  The new link is a small question mark icon / button on the “New Pill” screen.  I did not include the help text in the flow diagram for space / legibility reasons, but it would be a simple page with static content giving an overview of an example pills for each pill type.

This came from my heuristic analysis.

2) Added additional instructions / checkbox to close pill container and advance to next pill compartment before ending app session.  This is on the screen next to the drawing of the pill container.

This update seems more achievable than actually having the app and pill container communicate to verify that the pill container was closed and advanced.

This came from my heuristic analysis.

3) Removed customization by day from time selection screen.  Put customization by day on a separate optional screen.  This is visible on the farthest left two screens in the flow diagram.  It was done to make the setup process simpler for the common “same pills every day” use case.

This came from my heuristic analysis and a paper prototype usability review.

4) Clarify button labels on screen where pill / time combinations are added.  The old design was ambiguous as to whether “More” meant “More times to take this pill” or “More pills”.  The new design uses clearer verbiage.

This came out of my heuristic analysis and a paper prototype usability review.

5) Add screen to view next pill session and mark it as complete without going through phone call / reminder process.  This allows the user to, for instance, take pills a few minutes early and not still have to go through the phone call reminder process.  it is intended for “power users”.

This came out of my heuristic analysis.

6) Add screen to view schedule by pill and edit individual pills’ schedules.  This makes it easier for the user to update his / her schedule without starting from scratch.

This came out of my heuristic analysis.

7) Use clock time labels instead of ambiguous verbiage like “morning” or “mid-day”.  A user was not sure which category 11:00 a.m. fell into.  The clock time labels will be used in the app and on the pill container (which will be sold with labels that can be placed onto each compartment).

This came from paper prototype review subjective feedback.

8) Add ability to add multiple times at once on the schedule configuration screen.  This is visible on the upper of the two screens on the farthest left.  It makes the process of adding times simpler and more intuitive for the user.

This came from a paper prototype usability review.

9) Add day/time for current pill taking session to all screens in the reminder / in-session flow.  This is so that the user is always 100% what pills they need to take.

This came from a paper prototype usability review.

10) Add quantity of pills per occurrence to the “view schedule” page.  Also, add a way to configure the quantity of pills per time slot.  The quantity has been made visible in both the “schedule configuration” and “pill session” flows on key pages.

This came from a paper prototype usability review.

[apq3uw]

Prototype Test #2

Solution design update

Note : this section is duplicated in the Prototype Test #1 post.

After reviewing feedback on my storyboards and talking to people in class, I’m moving forward with a solution that combines the “pill container guarantees that only one pill is taken” aspect of the “Pill Master Pro” solution and the “system reminds you that it’s time to take pills” aspect of the “Pillbot”.  Since the Pillbot’s “the device dispenses pills for you” feature seemed highly problematic (breakage, jamming, wrong pills in a dispenser), that has been taken out of scope.  Instead, a mobile app will provide some of the reminder / scheduling features previously envisioned for the pillbot.

The proposed solution is: * A circular pill container that guarantees that only one pill container can be opened at a time, with buttons to allow the user to “queue up” the right container before the next pill-taking session. * A mobile app that is used to configure a pill-taking schedule - types of pills, names of pills, and dates / times when each pill is taken.  The app also makes a reminder phone call to the user at the time of each pill-taking session.   In the call, the app reminds the user of which pills to take and asks the user to confirm that pills were taken. ——————– Observations from the second Paper Prototype test:

* The “More” button to add additional occurrences of the same pill is weird / unnatural.  That button should be closer to where the time is entered so that all the time entry is clustered together.

* Different schedules by day confuse the interface and are not needed.

* The app design is oriented towards ongoing pill use but people sometimes have one-time or short-term pill needs as well.

* Pill container has morning / midday / night labels but application allows you to configure exact times.  This can lead to confusion - for example, it’s unclear whether 11:00 is “morning” or “midday”.

* Put the time for the current session on the “It’s time to take [pills]” screen when a reminder call is issued.

* A random photo from the user’s phone would work just as well as pre-configured “favorite photos” and would remove a configuration step (not shown in the paper prototype).

Additional feedback

* It’s important (though difficult) to get the focus exactly right in the scheduling interface.  Too much flexibility (unneeded allowances) is confusing and detracts from the important things that the user cares about.  Feedback from this user was that being able to add multiple times at once is more important than being able to customize times by day of week and that customizing by day of week may not be needed at all by many users.

* The “short-term” or “one-time” pill scenario should either be better supported or explicitly not supported.

* Getting the right pill is easy.  Customized / printable labels from the app would help address the “11:00 a.m. could be either morning or mid-day” problem.

* Being able to add / remove compartments from the pill container would also reduce possible confusion about whether a compartment is the “right” one at any given time.

[apq3uw]

Prototype Test #1

Solution design update

After reviewing feedback on my storyboards and talking to people in class, I’m moving forward with a solution that combines the “pill container guarantees that only one pill is taken” aspect of the “Pill Master Pro” solution and the “system reminds you that it’s time to take pills” aspect of the “Pillbot”.  Since the Pillbot’s “the device dispenses pills for you” feature seemed highly problematic (breakage, jamming, wrong pills in a dispenser), that has been taken out of scope.  Instead, a mobile app will provide some of the reminder / scheduling features previously envisioned for the pillbot.

The proposed solution is:

* A circular pill container that guarantees that only one pill container can be opened at a time, with buttons to allow the user to “queue up” the right container before the next pill-taking session.

* A mobile app that is used to configure a pill-taking schedule - types of pills, names of pills, and dates / times when each pill is taken.  The app also makes a reminder phone call to the user at the time of each pill-taking session.   In the call, the app reminds the user of which pills to take and asks the user to confirm that pills were taken.

--------------------

Observations from the first Paper Prototype test:

* User needs to be able to configure a quantity of pills per occurrence in the mobile app.

* List quantities as part of any schedule listing.

* User not sure of what to enter in “pill name” field (supposed to be user-understandable name for pill).

* Values in pill type dropdown unclear, could overlap - tripped up user.

* Pill type list missing pill types.

* Good: Time choices instead of manual entry is good - limits unnecessary flexibility.

* Good: Open button makes it clear what container will be opened.

* Would be nice if the phone confirmation of “pills taken” also “clicked” the “Next” button on the pill container rather than the user having to do that separately.

* 15 minutes to configure / review schedule of pills.

* Reminders should be issued every 15 minutes after the initial “time to take pills” reminder.  Better yet, let the user configure that.

Additional feedback

* The circle design of the pill container is effective.  It reduces ambiguity on which pills should be taken.  Fits with natural time flow.

* Having a fixed amount of compartments on the pill container is problematic.  You need more than one per day, but how many?  Current design has 3 per day but that could be too few or too many.

* The confirmation “reward” photo is good - could a game be used as well?  User responded well to the “gamification to earn the reward of a donation to a favorite cause” which I found in the “inspiration” mobile app I used for this assignment.

Conclusions

The pill container design works well.  Some way to customize the number of pill containers would be useful.

The mobile app works fairly well.  Pill quantities are an obvious needed improvement.  The overall experience was short of “this was really easy”; I should look to make general improvements.

[apq3uw]

Storyboard Feedback Interview #2

This interview was with the same subjects as “Insights from Interview #1″- a married couple.  The woman in her 60s and a man in his late 50s.  They live in a single-family house in a rural area.  Both are retired.

Storyboard #1: Basic pill management solution

Feedback

* Would be useful to be able to sense whether the container for a given day is empty.  There could be a weight sensor or a tiny camera within the container.

* Scheduling pill intake throughout a day is a problem not currently addressed by this solution.

* Color is important to emphasize which day is “active” for removing pills from the device.

* Confirmation of taking pills important.  There could be a button for the user to press to confirm that they have taken a set of removed pills, with the system raising some sort of alarm if it not pressed after a while.

* Having a solution not require a trained medical person to come and assist would be great as such people are very expensive.

Storyboard #2: Complex pill management solution with more features

Feedback

* Confidentiality is a concern with such a “smart” device.

* The “Pillbot” could easily be overheard by someone else in the house, or perhaps broken into (lots of pills inside).

* Seems like this could get quite large - not portable and could be unwieldy.

* Pill hoppers need to remain clean / sterile.

* What is a pill breaks?  What if the delivery mechanism jams?

* Auto-calibration to determine number of pills in a hopper (based on size or weight?) would be handy.

* Pharmacies / prescription drug plans are already doing a lot (especially recently) with automated refills.  This may not be an important feature for the pill manager.

[apq3uw]

Storyboard Feedback Interview #1

This interview was with the same subjects as “Insights from Interview #1″- a man in his late 60s and a woman in her late 50s that are in a long-term relationship and live together.  Both are still employed.  The man takes pills twice daily.

Storyboard #1: Basic pill management solution

Feedback

* Manual loading of pills is difficult and error-prone.  Mistakes are frequent even by competent adults.

* Taking pills more than once a day is a very common scenario.  A solution needs to support more than one “access” of the pills per day.

* One idea to address the “manual loading of pills is tough” problem is to have pills loaded at a pharmacy.  The pharmacy could fill up a pill manager (would need to hold more than one week’s worth) and give / /mail it to the consumers, who would later return the empty device.  Liability could be a concern; pharmacies may not want to be in the business of dividing up pills.

* In general, people can spend a lot (”way too much”) time interacting with their pharmacy - reducing that would be good.

Storyboard #2: Complex pill management solution with more features

* Good idea and would address some of the “manual loading” issues described above...but only if it is exactly right every time!  Lots of possibly points of mechanical failure.

* The digital calendar concept is a good one to help with forgetting.

* It would be useful to be able to be able to track whether a set of pills were removed from the “pillbot”.  One idea is to require lifting a “door” to access the dispensed pills.

* This solution would not solve all problems for all patients, but no solution will.  Some people are always going to need help.

* Somehow keeping the original bottle of pills associated to the “hopper” that holds the pills would be helpful.  Good visual indicator, could also assist with avoiding bad refills (into the wrong hopper).

* Visibility into the contents of each hopper could be useful.

[apq3uw]

Problem Statement Refinements: Week 3

Here are the refinements I made to my conception of the problem while working through the worksheet.  Not all of these would make it into a concise problem statement but all of them could influence solution design.

Refinement: Consider portability in the design.

Description: This came up during the persona analysis as I imagined the personae traveling but still needing pill management.  This concern points us towards a simpler / smaller / lightweight solution.

Refinement: Aim for a monitoring option other than video.

Description: This came up during persona analysis as I don’t see either of my personae wanting to use video monitoring of a daily task like taking pills.  The problem is finding another workable technical option (not that video monitoring is obviously workable, either, though.)

Refinement: Consider security / accessibility of pills.  

Description: This came up during the stakeholder analysis as I realized that small children and pets (who shouldn’t get into pills) are a stakeholder.  It also came up during my in-class critique.

Refinement: Consider medical personnel (e.g. user’s doctor) as a stakeholder.

Description: This came up during stakeholder analysis as I wondered what a doctor might think of each solution option.  The doctor may need to buy into / provide support for a user’s adoption of any solution.

Refinement: Consider keeping and displaying a count of pills remaining, even in a “simple” solution.

Description: This came up during the how / why ladder after realizing that knowing a count of pills was very important and could be done via simply displaying the information to the user even without systematically trying to use that information for automated reordering.

Refinement: Possibly de-prioritize the function of providing a reminder to take pills.

Description: This came up during hierarchical task modeling when I realized that having a reminder is important, but there are many other ways (e.g. smartphone alarms, clock alarms) to set up reminders using commonly-used tools.

Refinement: Consider non-pill uses of product.

Description: This came up during user journey analysis.  I realized that part of becoming a power user of any solution could be discovering and/or applying uses of the solution other than pill management.

[apq3uw]

Brainstorming Session

Here is a photo of the results of my brainstorming session.

Top three solutions:

1) Pill box that alarms if pills not taken.  Both of my interviewee pairs mentions pills as an ongoing issue.  Pill schedules can be complex and are not necessarily something people have had to manage throughout their lives.  It can be very easy to forget to take pills on any given day.  This “smart pillbox” feature would alert the resident, a designated caretaker, or both, when pills are not taken by a certain preset time each day.

2) Pill box that automatically reorders pills.  (This could be another feature of the “smart pillbox” described in #1).  The pill box would track supplies of each pill, order new pills when needed, and would create an appropriate alert when a reorder is not possible (e.g. when a prescription needs to be re-authorized by a pharmacist).

Those two solutions address two themes from the interviews: the need for frequent medical monitoring, difficulty with complex mental tasks, and poor memory.  Its specific problem statement is something like “People in their 70s / 80s need to take the correct pills at the same time each day.”  

The following solution addresses a different problem statement - something like “People in their 70s / 80s need to become aware of areas of their home that present risks of accidents.”  I’m keeping this very different idea around in case the “smart pillbox” concept becomes unattractive for any reason.

3) A robot that can explore a dwelling, locate “trouble spots” for accidents using vision, object identification, and analysis of how objects work well (or do not work well) together in a household environment.  Ideally, the robot would also be able to suggest remediations (e.g. move this piece of furniture away from the wall, put up a bannister here, etc.)

[apq3uw]

Insights From Interview #2

This interview had two subjects - a married couple.  The woman in her 60s and a man in his late 50s.  They live in a single-family house in a rural area.  Both are retired.

Questions and summarized answers

Do you think about a day in the future when you may have trouble living in your current home? - Yes.  When we designed the home, we explicity discussed "aging in place" with the architect. -- Layout. -- First floor master, laundry room bordering bedroom - possibility of one-floor living. -- Walk in shower with a seat. -- Plan for the future. -- Enough room for a wheelchair. -- No wood fireplace (gas instead). -- Elevated oven. -- Whole house generator was added after a scary power outage.

My last interviewees talked about the importance of personal connections.  How do you plan to maintain those? -- Friends group with scheduled activities (meals) -- Other scheduled activities -- Family connections very important -- Group workout class with trainer and group of women -- Historical society -- Crossword puzzles (every day) -- Limit TV (only set hours per day) -- No “normal” social media (yet)...though they do use an app to share photos of family members (grandkids)

Have you had family members who have had to move out of their homes due to complications related to old age?

- Mother of the male interviewee -- Her partner had a stroke. -- Meal preparation a real challenge.  Other challanges --- Socialization --- Cleaning --- More people aware of her location and safety --- Had a bad fall, needed a lot of help afterwards

What are some things that you enjoy about owning and living in your own home?

- View, rural nature. - Previously lived in an apartment (condo) setting with shared walls.  Difficult to fully block that out / pay no attention.

What are some things you worry about not being able to do in the future? - Be alone.  "Your business is everyone else's business" - the flip side of easy socialization at senior living facilities.

How have you observed, or maybe even helped, older family members stay in or move out of their homes?

- The male interviewee helped his mother stay in her own home / condo via performing some of the following tasks on periodic visits: -- Changing light bulbs -- Accessing storage space -- Seasonal changes - bring certain things out, put other things away -- Some cleaning tasks -- Strength, height reaches -- Shopping for specifically challenging items -- Appliance setup, new technologies

More on technology... - Resistance to new technologies - Guarding against resistance to change - Staying open...but maintaining that sense of self. -- Attitude towards technology will change - younger people are so much more used to technology in general - but some (many?) people will keep the sense of “I don’t want anything that wasn’t around when I was xyz years old”.

What would you envision being helpful to allow you to stay in your home in, say, 15 to 20 years? - Lessons learned from family members (aunt / uncle) who are still “at home” in their 80s - Starting to have things done for them by others. -- Cleaning service. -- Shoveling snow. -- Lawnmowing. -- Grocery delivery.

- Specific to their current home...

--- Someone / something to help with water filters for the well water --- Lifting 40lb bags of salt --- Purchasing, bringing home, putting into tank... ---- Some options to make the filtration less frequent, but tradeoffs (pricy)

- This could be relevant to anyone on well water (still a lot of people) --- Improved light bulbs (promise of LEDs)

Let's consider a few different ways to help people stay in their homes.

* Robot to perform tasks * Physical design of home / appliances / objects to make them easier to use. * Service to connect you with people who can come and perform tasks

Which of those sounds most appealing?  Why? - Service or appliances.  "If you're going to have a cleaning person, you don't need a cleaning robot." - Burners being left on stove - very big deal! -- Multiple stories of near-misses on big fires from acquaintances -- One family used a "baby monitor" - style camera to monitor an elderly person living alone --- One person left in the house is a HUGE difference from a couple. --- Lots of promise around designing a series of sensors to detect that something is wrong. -- Two people together can catch each other after mistakes.

- Fall avoidance very important - Design: avoid narrow openings, other things that can easily cause trips / falls.  - A robot could evaluate a house, find possible hazards.  The robot could watch resident walk around, detect possibly dangerous behavior.

What do you think of a “smart pill box”?

- Good idea. -- Very common to have to take a lot of pills -- Even people without prescriptions can take a lot of vitamins, etc. -- Simple boxes to sort pills --- Some pills require handling (e.g. cutting in half) --- Very easy to get off-cycle - "did I take that or not?"  Can easily lead to problems - do you take a catch-up dose? -- Could even include automatic reordering. -- No travel to pharmacy, or ideally even worrying about reordering.

Other challenges...

- Losing things -- Where is my purse?  Phone?  etc. --- Some sort of "thing finder"

Potential for autonomous vehicles to help 

- Autonomous vehicles -- Don't have to worry about car maintenance -- Take person to appointments -- Less reliance on another person coming to drive --- Easier for spontaneous outings -- Ability to get around without being able to drive safely -- Mobility could really improve despite the current model of "independence" changing

If you were to design the ideal tool or set of tools for helping you or your loved ones stay in their homes, what would they be? - Help with meals -- Lifting, timing, planning, loss of dexterity, opening containers - That nightly dinner is just too much - Important to get fresh food every day -- Improving quality of precooked meals - improved freezing / quality?? - Meals are a very big part of the day!

What are some possible problems / concerns? - Freshness - Community -- Can CSA-type businesses provide some sort of "fresh meals" service? - How to avoid having eating become a chore?

Are there any other thoughts about staying put while aging that you would like to share?

- This will become more and more important as the baby boomer generation ages.

These photographs illustrate possible “problem spaces” which could be problematic for an elderly person.  (Potentially, a robot could identify these for remediation or they could be key spots for monitoring if they cannot be avoided.)

This staircase, with its sharp turn with a gap in handholds, would be challenging to navigate.

This narrow space, used to access a lamp’s on-off button, would be difficult to get into (narrowness and need to bend over) and could easily result in falls.

Insights

* There are a lot of services available to support aging in place and they can be very effective.

* Two people aging in place together is very different from one person aging alone (repeat of Interview #1)

* No longer wanting to take on challenges / go out in public is a bad sign, particularly in someone who wanted to do such things earlier in life.  Things may go downhill fast after that.

* A single task (e.g. the well filtration system example described above) can be a huge deal for people aging in place.

* Wide spaces are important to allow people to get around with bulky equipment like wheelchairs.  (They may be helpful to robots, too!)

* Meals are tremendously important!  When done well, they can be a big quality-of-life enhancement.  There is a lot to think about - mental agility (meal planning), strength, dexterity, coordination of multiple tasks, freshness of food, possible companionship, use of appliances...

* Issues with forgetfulness can quickly compound - for example, even after you realize you’ve forgotten something about your pills, you have to decide whether to take the “forgotten” pill, etc.

* Pill management is more than just remember to take a pill on any given day.  Need to manage many pills at once, reorder when needed, possibly manipulate pills (e.g. cut in half).

* Autonomous vehicles can help people run errands and socialize without needing to be able to own / maintain / operate their own vehicle.

* Getting people to maintain an open-minded attitude about technology may never stop being a tough nut to crack.

Problem Refinements

* Forgetfulness is a big problem.  It seems that a lot of people who age in place (ostensibly successfully) “skate by” through near misses.  Even younger people can easily be forgetful.

* The “smart pill box” idea addresses both forgetfulness and health.

* The user group for a “smart pill box” includes people who take vitamins / minerals as well as people on medication (of course, many people take both).

* Mental engagement / problem solving / talking to other people (e.g. not “shutting down”) is a nice-to-have for any tool that assists aging in place.

[apq3uw]

Insights from Interview #1

This interview had two subjects - a man in his late 60s and a woman in her late 50s that are in a long-term relationship and live together.  Both are still employed.  They live in a large house on many acres at the end of a dead-end street.  The man works out of this house as a small business loan consultant and the woman is a schoolteacher.  They have no immediate plans to leave their home but acknowledge that it is not the sort of place “old people” tend to live.

Both subjects had a lot to say and the discussion flowed naturally.  In particular, there was a lot of wide-ranging discussion in response to the first question listed below.

Questions and summarized answers

Do you think about a day in the future when you may have trouble living in your current home?

- Yes.  Concerned because of size of house.  Many general maintenance tasks need to be done on a regular basis. -- There are also "Bad things" that can happen - furnace out, power out, anything that would require sudden movement and cannot be planned for in advance.  Perhaps more concern about things that require sudden action than about things are known in advance. -- Difficult for a person who has always done things himself to let go of tasks like plowing, cutting the lawn. -- Ladders / anything requiring more than one step up.The interview covered a lot of ground and some of it was quite abstract.  One of the most concrete / physical areas of discussion was household appliance.

-- A simple fix for a lot of stuff now would be to make sure that all technology in a current home is tied into a monitoring system.  Addresses the "I've fallen and I can't get up" problem.  If the technology could determine that you haven't moved in a long time and raise an alarm accordingly, that would be very helpful.

-- Affordability very important - smart houses are currently too expensive for many.

Have you had family members who have had to move out of their homes due to complications related to old age? - The mother of one of the subjects did.  She moved from a normal SFH to a detached house on the grounds of on assisted living facility to an apartment within the main building of the facility and eventually into a a smaller apartment. -- The surviving spouse in a married pair may not have a history of doing all tasks e.g. doesn't cook, doesn't work outside, etc.  So, doesn't want to perform certain tasks even if is physically able to.

-- For villa to apartment move, socialization was a big factor.  Was isolated in detached house.  Difficulty with walking to main facility for social opportunities (i.e. meals).

-- Particularly with one surviving spouse (and possibly with people who were always single) - isolation becomes VERY important --- Technology can enable interactions e.g. Skype, etc. --- Need to be able to participate in some sort of thought-provoking activity, otherwise brain capabilities deteriorate.

---- Need something that interacts with the person - makes them think!  Continue problem-solving throughout life to keep brain working.  Need sense of accomplishement - "I am necessary".  Sense of self is based around what is accomplished.  Could virtual reality help with this?

- Health can change very rapidly in an old person. -- Monitoring / rediagnoses very important. -- Some things already monitorable. --- Need more things - e.g. a diabetic requires monitoring of glucose. -- Confusion about which medications to take and when.

Let's consider a few different ways to help people stay in their homes.

* Robot to perform tasks * Physical design of home / appliances / objects to make them easier to use. - Central vac would be very helpful. - Needs to be simple, controllable, has limits and avoids becoming dangerous (e.g. a stove that turns itself off) - Washer / dryer in single unit; all done as one interaction. - Safety is very important!! - Difficult to change sheets. -- Touches on the key area of personal hygiene.

* Service to connect you with people who can come and perform tasks - The people doing that service need to be part of the monitoring / caregiving process. - Check on the resident along with performing the task (e.g. doing the lawnmowing). - People need to become part of the process. - If you want to be a tradesman who serves the elderly, you need to be a caregiver as well. - In assisted living, the person who cuts the grass is not the same person who does health monitoring.  For aging in place, that needs to be the same person. -- "It takes a village" concept. -- Interaction, visit.

Which of those sounds most important?  Why? - (one subject) Medical monitoring and tracking of changes.  Need very rapid detection of any deterioration of health. - (other subject) Socialization instead of isolation.  An isolated person can have very low quality of life.  Need ways to keep people engaged with other people. -- People need to be able to use modern technology - even if it would normally be hard for them to learn - technology literacy.

--- Unclear if modern generations are sufficiently tech-savvy to remain “with it” for life or if we will always have the problem of people struggling / not wanting to learn things that were not part of their childhood / middle age.

Below are some appliances that were called as problematic for the elderly during the interview. 

This gas stove could remain on indefinitely.  It is also possible to turn the gas on without actually lighting the stove (much worse).

The immersion blender was cited as an appliance that could easily “get out of control” and cause injury.

Using a paired washer and dryer involves a lot of lifting and/or bending over.  A unit that either washed and dried clothes in a single structure or somehow automatically transferred clothes between the washer and dryer would reduce the amount of challenging work.

Insights

* Even if you successfully account for planned home maintenance tasks, unexpected events (power outage, appliance breaks, etc.) can still be a big challenge for elderly home residents.

* The problem of aging in place is very different for two people living together than for someone who has lost their spouse / companion.  (It is surely different still for a long-term single person.)  When two people live together, they naturally acquire different areas of expertise.  One person by his/her-self may not be able or want to do all the tasks required to age in place after the other person has passed away.

* There is a lot of potential in solutions that are “integrated” in non-obvious ways.  For example, rather than having separate people mow the lawn and provide in-home checkups, have the lawnmowing person also come and check up on the resident(s).  There may not be people with the required skillsets now but they could develop in the future.

* From a medical perspective, constant monitoring is extremely important.  Health can change very quickly and early detection is paramount.

* The major concern with appliances (and other things) is absentmindedness / forgetfulness e.g. forgetting to lock the door, forgetting that the stove is on, forgetting which pills to take.  Appliances should be designed to detect and counteract this sort of behavior (e.g. a stove that turns itself off after a while).

* Aging in place requires continual challenges to the mind.  Daily problem-solving and challenges are paramount.  

Problem refinements

* The problem of aging in place is a big one and my current problem space is broad and needs to narrow.  Out of the several possible areas listed above, this interview left me most intrigued by 1) smarter appliances, 2) constant health monitoring, and 3) creating mental challenges.

* Somewhat surprisingly, my original idea of “create robot to perform specific tasks” didn’t get much traction.

* Right now, I could see pursuing solutions related to appliances preventing bad things from happening - for instance, a smart pillbox that only dispenses the pills for the appropriate day of the week.  I could also see pursuing an “your appliances collectively monitor your health” solution.

[apq3uw]

Project Idea #3

Edit 1/14: This is the project idea I’m going with.

User group: Aging people who want to stay in their homes / apartments but may need to move to an assisted living facility

Task / activity: Interact with a robot that will be able to take care of necessity daily and home-maintenance tasks that the resident(s) need to to in order live independently but cannot no longer perform.

Current solution: None that I know.

Challenges: I think the biggest challenge is the variety of tasks this might entail.  Some people may no longer be able to use a stepstool to change light bulbs.  Others may not be able to bend over to sweep a floor.  Climbing stairs, operating appliances, and taking care of the exterior / grounds of a property are more of the many tasks that a robot may need to perform in order to let people live independently.  It will be a challenge not only to create robot(s) to perform all of these tasks but also to let the resident(s) tell the robot which task(s) need to be performed at any given time.

[apq3uw]

Project Idea #2

User group: Anyone who enjoys card games such as hearts or spades and could learn to play bridge

Task / activity: Learn to play the card game bridge at a novice level

Current solutions: There are many books and in-person teaching guides.  There are starting to be some interactive / online teaching tools such as http://www.learn2playbridge.com/ .

Challenges: The learning curve!  Bridge is a very complicated game and most people need to spend a lot of time learning before being able to play well.  There are many more leisure activities available now than in the game’s hey-day in the middle of the 20th century.  There is also competition from poker.

A good bridge teaching tool will convey the richness and fun of the game to the learner without “swamping” him/her with too much information or otherwise turning him/her away from the game.

[apq3uw]

Project Idea #1

User group: Children (ages 8-10)

Task / activity: Control a computer or a machine like a robot.  This could be via writing code, using visual tools that result in code, or controlling a machine like a robot.  This will teach the child about the computers that are already so important to his/her life and give them a basic understanding of computer science to build on later in life if desired.

Current solutions: I know there are programs that attempt to give many children a taste of computer science.  One specific one that I am aware of is www.code.org

Challenges: Attention span.  I am concerned both with the students not having any interest in the task and with them having too much interest and wanting to do programming activities to the detriment of things like playing outside.

[apq3uw]

About Me

My name is Andrew Quinton.  I live in South Seattle (Columbia City) with my wife and one-year old son.

I grew up in Maine and went to Swarthmore College in Pennsylvania, where I majored in Math/Stat and minored in Environmental Studies.  PMP is my only post-graduate work and this is my seventh course.

I have worked at Liberty Mutual Insurance since 2008, first in New Hampshire and here in Seattle since 2013.  At Liberty, I’m a team lead in a group doing back-end service development (mainly Java), primarily for applications used to sell and service auto and property insurance policies.  These days, I do more reading than writing of code.  I consider myself more of a generalist / problem-solver than a true “coder”.

I have little experience in HCI / Design.  I designed the UI for one application for which I was the only developer but can’t say I put much time / effort into it.  At my day job, I sometimes participate in discussions such as:

1) How to balance “keep the application simple” with “gather all desired information”?

2) How to only gather information actually needed for this particular user?

These discussions get me to put myself into the users’ shoes, but they’re a small part of my job.

Outside of work / PMP, I do normal “parent of small child” stuff - exploring our neighborhood, trips to the park and coffee shops and trying to get enough sleep.  When I had more free time, I did more substantial hiking, played bridge (the card game), and helped compile statistics for college basketball games (most recently at Seattle U).  I remain a baseball fanatic!