Data collection for AWS IoT Analytics
AWS IoT Analytics is a web-based application that helps collect, process, store, and analyze data from IoT devices. To explore more about this service, check out AWS News Blog.
Overview
AWS team collaborated with Teague, one of the leading design agencies in Seattle, to create the first, preview version of the product.
Key challenges that we were aiming to tackle:
Find the common concepts of collecting and processing data.
Create a robust experience for data engineers and data analysts
Create guiding experiences for less technical users.
Business goal
Create a streamlined experience for collecting and managing data for IoT analytics, flexible enough to satisfy the needs of users with a different familiarity of the field.
Role and duration
Lead designer for several areas of the product. I was part of the design team at Teague and closely collaborated with AWS development and design team.
August 2017 - October 2017
Research
The research phase for this project included:
Stakeholder interviews;
Competitor research;
User research and personas.
Meeting with key AWS stakeholders helped us understand the business and technical challenges. Together we were identifying users' main goals, addressing their pain-points, and discussing the possible solutions. The application is meant to be used by Data Analysts, who are responsible for analyzing IoT data, and Data Engineers, who are setting up the data channels and configuring the process of data collection. However, we also kept in mind, that not all businesses might have Data Engineers alongside Data Analysts, so we tried to minimize the technical complexity of the application by providing logical sequence, simple setup flows, guiding language, and minimum inputs that require programming knowledge.
Task flows and information architecture
The information architecture of the application is based on the users' task flows and their most common use cases. However, despite the tool will be used in chronological order, such as collecting, processing, storing, and then analyzing data, we decided to prioritize Analyze section above everything else because we expect that 80% of the time users will spend there, and it's the main reason why they'd use AWS IoT Analyze.
Ideation process: understanding the product flow.
Mental models
One of the main challenges was to understand users' mental models on how they move through the process of collecting and analyzing data. We explored multiple approaches to solve this problem, keeping in mind most possible scenarios as well as technical and time restrictions. We decided to keep each step of the process as simple as possible and separate from one another, however, we still provided guidance for the users to move smoothly from one section to another by showing "toasts" with the next steps and having the functionality to link different phases of the flow into one.
Ideation
Wireframing helped our team keep the process solid and simple and explore complex task flows and concepts without wasting too much time on every detail. Almost everyone in both teams was involved in this process, and we could fast and easily communicate our ideas or even technical limitations.
Prototyping
Our team was creating, presenting, and iterating low-fidelity flows for different scenarios. Tackling the main task flows helped us think about the product as a tool for action rather than static screens and states, and spot the gaps early on any inconsistencies in design patterns.
Validation and iteration
At every phase of the process, we worked closely with the AWS engineers, designers, researchers, and product managers to explore and validate our ideas and findings. Well-established communication between two teams played a crucial role in successful product design and development. We were able to respond quickly to changing requirements and iterate and make sure we meet users' and business needs, as well as deadlines.
Result
One of the main challenges of this project was creating a seamless experience for enterprise products for different task flows and with complex functionalities. We reached this goal by exploring different mental models, use case scenarios, a lot of iteration, and the unification of different components with visual design and animations. We created our final deliverable both in InVision and Axure to showcase interactions and animations for different components of the system. We used and expanded the AWS IoT style guide to keep the new product consistent with the rest of the IoT services.
The product was first released as a limited preview during the AWS conference re:Invent 2017, and now it’s generally available. To find out more, take a look at what TechCrunch says about it.
