Milestone III: The Design

Summary

MultiManager is an intelligent mobile app designed for persons who are always on the go and habitually check their phones for new information. The system proactively serves its users content from their most used apps, in a single interface. Consider the app an assistant. Users will have the ability to control which app categories they want to feature as well as choose their personal assistant — changes the visual design and voice of the app.  How people "use" their apps will equate to tasks that will be organized in up 10 categories that may include news, finance, shopping, entertainment, travel, utilities and games. The system’s can-do attitude will leverage historical and real-time data, and advanced algorithms to predict and mimic the owner’s future phone use. In addition, the system will define favorite apps based on metrics such as session length, session interval, frequency of use, screenflow and retention.



Design Space

With any project there are compromises. An earlier concept I explored was to use unconventional vs conventional UI patterns. As a designer, we are taught to think outside of the box with fresh ideas. Using existing UI patterns felt like I was taking a shortcut. I took comfort in the fact that there was no reason to reinvent the wheel when there are UI patterns already tried and tested. Users have mental models from experiencing other interfaces so when they commit an action it should come close to meeting their expectations.


Another tradeoff I had to consider was quantity vs quality. I wanted to give users an experience that felt like artificial intelligence with all of the bells and whistles. Toward the end of my competitive market research for similar products to the MultiManager concept I found that my app fell into a category called virtual personal assistants. I came across a Google interface called Google Assistant. The app boasts a component that gives users the right information at the right time. Feeling a bit discouraged because of the parallel ideas, I wanted my app to do even more. Unfortunately, giving users too many options can lead to feature bloat, which my lead to frustration, cognitive overload and ultimately abandonment. I resorted back to keeping the UI simple.


An even harder requirement to support was finding the best solution for bringing content to users who have formed habits of searching for information themselves. “It turns out that every habit starts with a psychological pattern called a "habit loop," which is a three-part process” (www.npr.org, 2012). First, there is a trigger such as a beep from the phone that tells the brain to switch to autopilot. The sound peeks the user’s curiosity, which leads to the second step in the loop — engaging the habit, resulting in the owner checking his/her phone. No real focus is needed to complete this action while in autopilot. The final part of the process is the reward. The emotions associated with closing the habit loop are feelings of being productive (79%) and happy (77%), according to a 2014 Pew Research Center survey. Other experiences participants noted from using a smartphone were frustration, distraction and even anger. This research tells me that although people feel good about the act of checking their phone for whatever the reason they are not necessarily being satisfied by the content they are consuming.


The Design

The Multi-Manager app will be skinned in a flat-design style. The other side of the tradeoff was to take a skeuomorphic design approach were the app could be textured in a leather binder for example; with lined, paper pages to resemble the daily planners people used to carry. I expect fewer interactions between the device and the user compared to the check-in habits that exist today.


Task One: User goes through the registration flow — signing in or creating an account — and lands on a page with options to sync their favorite apps automatically or manually (Figure A).

Figure A: Task 1, syncing apps automatically via the system.


When synchronization is complete, the system will display 10 recommended categories such as entertainment, communication, games, finance and navigation (Figure B). In this sketch, I planned for the category cards to flip upon, showing the user’s favorite apps associated.

Figure B: Task 1, categories/apps recommended by the Multi-Manager system




Contrarily, the smartphone owner can sync their personal apps with the system manually also. This gives them control over the apps they want the Personal Assistant to manage, however, it overrides the system's ability to provide the optimum experience which is based on app usage (Figure C).


Figure C: Task 1, syncing apps manually via user selection.


Task Two: Users can create their own category cards (Figure D). The user does this by selecting from a predetermined list of categories then the system will sync their “most used” apps that align to the newly created cards. These options give different user types choices for customizing their own experience.  


Confirmation messages such as, “Calibrating…”, “...successful…” and “...complete.” are shown throughout the journey to give users feedback of assurance.


Figure D: Task 2, creating a card — task category — for system synchronization.

Task Three: Takes place over a 24-hour period inside the Personal Assistant (PA), listing tasks labeled as categories (Figure E). As the day progresses, the PA assigns the user’s favorite apps to task categories. The categories constantly jockey for position, swapping positions up and down based on user data. The user can expand a category within the accordion design pattern to drill down a layer to see the apps associated. Single lines of copy then give a preview of the content inside, scrolling from right to left like a ticker tape. Tasks can be easily deleted by swiping the row to the right off screen.

Figure E: Task 3, viewing and modifying the daily calendar.

Future Technologies and Social Implications

The MultiManager app at its core is under the umbrella of artificial intelligence (AI) but would be considered by some experts as weak AI, which is computer technology built to perform a set of tasks. In contrast, strong AI is technology that comes close or exceeds human cognitive abilities. There are plenty of examples of weak AI that are system based and developed to perform specific tasks such as video games, credit card fraud detection and Deep Blue who defeated the chess champion Garry Kasparov. AI that is less system based and more robotic are steadily being developed as well such as Roomba the robotic vacuum, the Mars rover, and autonomous vehicles. However, scientists are continuing to pursue developing strong AI that is robotic and takes the shape of a human with common sense. Robotic technology today can certainly solve complex problems that surpass the human brain but may not be able to make some simple distinctions of that of a child.


Take Nadine for example, just one of many human-like robots (Figure F). She has soft, skin-like texture, can smile, make eye contact and shake hands. She works as a receptionist at a university and can greet visitors. Nadine can recognize previous guests and have conversations based on previous communication. She was programmed to have a personality, mood and emotions. Although there are still limits, advancements in AI is evident.


Figure F: Nadine with creator Nadia Thallman. Screenshot from telegraph.co.uk.

There are social implications surrounding AI. One concern is that people do not like robotic technology that looks like them — humans. According to a study that focused on social robots, participants said these robots threaten the uniqueness and identity of humans, “the more the robot’s appearance resembles that of a real person, the more the boundaries between humans and machines are perceived to be blurred (spectrum.ieee.org, 2016).” People also feel threatened by technology that can do a task better or ultimately replace them.


The proponents of AI say that robotic technology is the natural progression of human intelligence. Whereas, the opponents urge that there be limitations.


If the MultiManager app were launched and became popular among users, the effort to reduce phone check-ins could have an opposite affect and induce new habits with unforeseen dependencies on mobile devices. Further research would be needed to discover what effect this would have on smartphone users.



REFERENCES​​​​​​​​​​​​​​