Making an invisible pollutant readable on any screen

Sensirion

Making an invisible pollutant readable on any screen

Sensirion asked for one icon for volatile organic compounds; the work became a research project and a 120-page framework

User Research / InterviewsPrototyping / ConceptsDesign StrategyFacilitation
Years
November - January 2023
Domains
Air quality monitoring
Team
1 Researcher, 2 Designers
Activities
Market Research, User Research, Visual Design, Usability Testing, Guidelines Development

Reading time: 2 min

All projects

01 · About

About

Sensirion, a leading manufacturer of environmental sensors, approached the team with what seemed like a straightforward request: create an icon for Volatile Organic Compounds (VOC) and develop guidelines for air quality monitor producers.

The real challenge ran deeper: VOC data is complex and unfamiliar to most users, and needed to be displayed on devices with varying screen sizes and technical limitations. Sensirion's sensors also primarily show relative data (changes in contamination levels) rather than absolute values, adding another layer of complexity.

The Sensirion SGP41 gas sensor the work was built around.

02 · Process

From alignment to prototyping in five stages

  • Alignment: business goals study, project roadmap definition, research methods definition
  • Learning: niche research, trend research, market research
  • Analysis: user research, use cases definition, visual research, recommendations
  • Ideation & Envisioning: hypothesis mapping, ideation sessions, brainstorming, ideas refinement
  • Design & Prototyping: design concepts, prototyping
  • Evaluation: user testing, feedback gathering, design update

03 · Discovery, the reframing

An icon request became a research project

Original Request

  • Create an icon for VOC
  • Develop basic guidelines for AQ monitor producers

Expanded Scope

  • Research VOC sensor visualization and user needs
  • Provide guidance on data visualization: icons, colors, and micro-interactions
  • Create visualization concepts for low-resolution screens
  • Develop a holistic design approach for air quality data

Outcome: the team transformed the initial request into a comprehensive research project.

04 · Research

Four personas and a device audit before any visualization

Niche Research

  • Analyzed popular air quality monitors
  • Clustered devices by screen size and capabilities
  • Identified technical constraints across device types
Comparative design review of air-quality devices, Amazon review analysis, and the device clustering.
The device audit. Every monitor on the market compared by screen type, value gradation, colour and sound, then clustered by how they show anything at all.

Trend and Market Research

  • Examined leading solutions for interaction patterns
  • Analyzed value display conventions
  • Studied scale representations, color usage, iconography, and sound indicators
  • Conducted desk research on market trends
  • Analyzed competitor approaches to air quality visualization
  • Identified gaps in current visualization methods
Market tendencies: HVAC market size, IAQ trends, and connected air-quality offerings.
Trend and market research: where indoor air quality was heading, and who was already selling it.

User Research

  • Defined key use cases and scenarios
  • Identified user mental models regarding air quality
  • Understood user expectations for air quality indicators

4

User personas

Analysis of Amazon comments and the four user personas that came out of it.
Amazon reviews became the raw material for the personas: what people actually complain about when they own one of these.

05 · Key insights

Nobody knows what VOC means

  • VOC Unfamiliarity: users had limited understanding of what VOC means, making any icon solution inherently difficult to recognize without education.
  • Relative vs. Absolute: Sensirion sensors primarily show relative data, not absolute values, creating challenges for straightforward visualization.
  • Device Diversity: air quality monitors range from simple LED indicators to full-color displays, requiring flexible visualization approaches.
  • Context Matters: users need context to interpret air quality data meaningfully; a number or color alone isn't sufficient.

06 · Visual research

How to color an invisible phenomenon

Conducted visual research on existing VOC data visualization approaches for color coding, iconography, and data visualization, identifying best practices and proposing more effective visual solutions.

Colour semantics: green for poison and toxic, yellow for warning, orange for strong warning, blue for water and air, purple for toxic and contaminated, red for restrictions and danger.

Color coding

  • Analyzed existing color schemes
  • Studied color perception across contexts
  • Developed recommendations for accessible color usage
Six icon metaphors for VOC: fumes, a benzene ring, a leaf, particles, a molecule, and an emission cloud.

Iconography

  • Evaluated existing visual solutions for invisible phenomena
  • Created variations of VOC icons based on different metaphors
  • Tested recognition and understanding with users
Two chart treatments: absolute values against each other, and a relative reading marked against its own average.

Data visualization

  • Developed concepts for displaying relative and absolute values
  • Created multiple visualization options for different screen sizes
  • Tested comprehension and preference with users

07 · Concept ideation

Every concept on the table, and 'Air is Good' won

After brainstorming sessions, iterated on all possible VOC visualization concepts, then chose and tested the best concepts with users, landing on "Air is Good."

One concept on a fixed segment display: PM2.5, temperature, humidity, and the tVOC trend reduced to an arrow.
  1. 01

    Research findings

  2. 02

    Ideation

    Voting

  3. 03

    Moderated user testing

    Feedback analysis

  4. 04

    Concept improvements

    Voting

  5. 05

    Unmoderated user testing

    Feedback analysis

Two rounds of VOC visualisation concepts: the first spread of options, then the refined set on dark.
Every concept on the board in the first round; the second narrowed to the relative-to-average readings that testing preferred.

08 · User testing

10 moderated sessions, then 52 respondents at scale

Two researchers running a moderated session on a laptop.Participant feedback cards beside the concepts they refer to.Moderated study

10

Sessions

We conducted 10 moderated testing sessions to gather qualitative feedback on our concepts.
Concept options from the unmoderated study, with respondents' comments beside each.Unmoderated study

62%

52

Respondents

Never thought about the VOC level in their homes.
We expanded testing with 52 respondents to validate our refined concepts at scale.

09 · Solution

Not an icon: a 120-page framework

Rather than delivering a single icon as initially requested, the team created a comprehensive 120-page report with guidelines and recommendations covering:

  • Data Visualization Framework
  • Iconography System
  • Color Usage Guidelines
  • Interaction Patterns
The report structure: twelve numbered sections and every page inside them.
The structure of the deliverable, section by section.

10 · Impact and results

From tactical icon design to strategic UX

  • Problem Identification: helped Sensirion understand the true complexity of VOC visualization and brought these insights to stakeholders.
  • Hypothesis Validation: validated key hypotheses about user understanding and visualization preferences.
  • Actionable Recommendations: provided a fully actionable plan with specific implementation guidelines.
  • Strategic Value: elevated the conversation from tactical icon design to strategic UX considerations.
The VOC Data Visualization Report open on a tablet, with its spreads laid out behind it.
The report itself: research objectives, the device review, colour, iconography, data visualisation, and the guidelines that came out of them.

11 · Conclusion

The true challenge was representing complex air data

What began as a simple request for an icon evolved into comprehensive research that addressed the true challenges of representing complex air quality data, delivering lasting value to Sensirion and their hardware partners.

Next project →

Building cloud infrastructure discovery tool