Our goals for the IoT design was to create an onboarding experience that reassures the safety of
users' private information and a voice UI that encourages casual conversations between humans and AI.
I was involved in all stages of the UX design process, but my main contribution focused heavily on creating the user flow and visual design direction.
Timeframe: September, 2020 - October, 2020
Tools: Miro, Discord, Figma
WHAT ARE THE EXISTING PROBLEMS?
Feedback we received from users across brands such as Google and Amazon,
expressed their concern toward their privacy and interactions with the AI systems.
They were worried that the devices were passively “listening” to household conversations and collecting
private information. In addition, voice AIs require specific commands to function properly; which makes them incapable
of answering vague requests.
An IoT onboarding experience that guarantees confidentiality of private information and a voice UI that is capable of understanding various utterances and speech.
We determined a Terms and Policies feature would solve the security problem. Users would gain trust with a product once they know their information is safe and won't be utilized by any third-party.
To align with our user needs, we deduced that text displays in the voice UI should appear less robotic, and more casual.
Our first step was to expand our knowledge on smart speakers. We researched companies such as Google and Sonos to understand their design practices for
IoT development. Once we touched base, we had our lead UX Researcher, Vic, conduct user interviews to dig deeper into the UX problems.
Our personas were centered around general necessities such as music and clear speaker quality, with the assurance that
their personal and household information are safe.
WIREFRAME AND USABILITY TESTING
After we spent some time developing an optimal onboarding experience, we tested our theory and here are the results from our testers.
5/5 - Liked the onboarding process and how it tells users that private information remains confidential. In addition, we received positive
feedback when our testers read through a prototype conversation between a human and the AI.
Within the onboarding experience, users will encounter two features that would guarantee the safety of their private information:
Smart ID and Terms and Policies.
Smart ID enables users to log-in to the app via their voice; which would help eliminate identity theft.
Our Terms and Policies was structured around a proprietary cloud system and information would only be used
to improve future devices, and would not be leaked to any third-party.
VOICE UI SOLUTION
We kept in mind to design the voice UI to "understand" various utterances and speech. We created a scenario where a user
speaks to the AI as if it were another human, and the AI gives a human-like response.
Overall, I learned that disconnections happen often as it disrupted the team's workflow and we had to reassess in the beginning.
However, we recovered quickly and I took the opportunity to learn about my team's strengths and weaknesses and offered to assist various tasks to expedite the workflow.