Duy Huynh Anh
Designer
Researcher
Maker
Designer
          Home

Pollen

2026 A mobile app to encourage citizen scientists to participate in pollinator monitoring within urban environments by removing the knowledge barrier while encouraging connectivity and engagement. 
Context Problem Space Urban land is projected to grow by 50–69% by 2050, with around 70% of the world's population living in cities. Scientists need continuous data on how insect populations change, and citizen science apps could help collect it. But today's apps, like iNaturalist and iRecord, are built around logging one sighting at a time through forms, and neither tracks abundance at one site over time. They feel made for scientists, and once you press submit, you never see what your data achieved.

Automated monitoring takes the data entry away, which leaves the app free to focus on what keeps people engaged. How might we remove the barriers to collecting insect data and give people an incentive to take part?

 Brief Design Pollen, a mobile app that turns automated pollinator data into something Gen Z users understand, trust and want to come back to. The app pairs with a small camera monitor that identifies insects, so users never have to log a sighting themselves. It's for young people who are trying to reconnect with nature and care about the environment enough to act, but don't have the time or expertise for traditional species recording.

 I triangulated three methods to see where current platforms fail, in both usability and motivation. Three participants in their early 20s logged a real sighting on iNaturalist or iRecord while thinking aloud. I then ran two semi-structured interviews with them about setup, feedback and community. I also studied two analogous systems: Google Home for low-effort device setup, and Forest bike share, where riders photograph a parked bike to confirm placement. That photo check became Pollen's placement step.

 Key Findings The current apps feel like a scientific form, and submitting leads nowhere. Forms and dense maps made logging feel like a survey, and every participant called the moment after submitting anticlimactic.

"This feels very much like it was made for a scientist." (Participant 1)

Personal trends are the hook. Participants compared it to screen time and fitness tracking, and wanted to see change over days and months.

"I really like being able to say, oh December 4th, there were 10.
But December 5th, there were none." (Participant 2)

People want to do it together, but not competitively. Both sessions independently asked to see contributors nearby. Light gamification such as streaks appealed, but "Pokémon Go" mechanics raised worries about people disturbing habitats.


 User Journey Mapping how someone logs a sighting today showed where motivation drains away. People start keen, but photographing, identifying and form-filling wear them down, and submitting leads nowhere. Each pain point became a design opportunity for Pollen.




Development From Insights to UX goals Journey mapping across the existing platforms and analogous products pointed to three priorities, each tied to what users said or did.

  • Transparency
    The post-submission "black box" was the most common frustration, so every dashboard shows how the user's data feeds into datasets and research papers.

  • Engagement
    Users wanted personal trends and to feel part of a community, so the app has trend charts, lifetime stats, a community map and shareable summary cards.

  • Learnability
    Participants would give up if getting started took more than 10–20 minutes, so the app uses familiar navigation, plain language and a guided seven-screen setup with one action per screen.


 Service Ecosystem Pollen is more than an app. It connects users at home, a camera monitor, a local community and the researchers who use the data. The monitor identifies pollinators with on-device machine learning, while a second model checks the camera is working so the app can flag problems before data quietly goes bad. Friends and neighbours appear through the map and shareable cards. Researchers use the continuous data in datasets and papers, and the app reports each contribution back to the user, closing the loop existing platforms leave open.


Three Core Flows With identification automated, the user's role shifts from data collector to informed explorer. The app is designed around the moments people come back for, with a short one-time setup to get there.

  • Get started
    A new user powers on their monitor, scans a QR code, follows one placement instruction per screen, then takes a photo to confirm placement. If pairing or placement fails, the app sends them back to fix just that step.


  • See your impact
    A returning user lands on a dashboard of trends across all their monitors, picks one, then opens individual sightings showing species, time and confidence. Every level carries a research contribution summary.

  • Monitor together
    The community map shows who is monitoring nearby. Users can browse the leaderboard and their profile, and share a summary card of species seen, days active and rarest find. If no one is nearby, the map zooms out to the national view and invites them to share the app.



Iterations: From Paper To Pixel
I conducted three rounds of iteration. Each round removed a decision users shouldn't have to make. I tested two paper prototypes with the same three participants as Wizard of Oz sessions, using a physical stand-in for the monitor device, then moved to Figma for three wireframe iterations.

Setup. Picking the monitor from a Bluetooth list confused everyone, and a bare loading spinner left two participants unsure what to do. I replaced the list with a QR scan and cut the loading screen.

Data and Social. Nobody noticed the carousel arrows and the leaderboard's rules were unclear. I swapped the carousel for a list andadded context to the map and gave the profile real stats.

Refinement. I added device naming, a Do/Don't placement screen and a sighting detail that shows research papers. I renamed the tabs Home, Devices and Map, added a week/month/year toggle and a Lifetime Impact panel.



First Iteration
Using AI As A Second Opinion Alongside my own wireframes, I used Claude to generate an independent set from the same brief and research. Both landed on the same setup structure and three-tab navigation, which gave me confidence in those decisions. The comparison also exposed weak tab labels, which led to the Home, Devices and Map naming.


First draft made by me


First draft made by Claude



Second IterationThe second iteration builds on the what I learned in the first round and combine both original and AI inputs.


 Third Iteration The third wireframe iteration focused exclusively on Flows 2 and 3, incorporating feedback from the previous round and introducing a small number of targeted A/B variations to resolve outstanding layout questions before moving to high fidelity.




 Outcome ApplicationPollen is a citizen science app for people who want to reconnect with nature without becoming experts. A small camera monitor at home identifies visiting pollinators automatically, and the app turns those sightings into something worth checking in on: trends over time, a gallery of real insect photos, a map of who else is monitoring nearby, and a clear view of the datasets and research papers your data has helped. Instead of logging sightings through forms, users set up their monitor once and watch their impact grow.




 Final Prototype The final prototype, built in Figma, covers the full experience: a guided seven-screen setup, a home dashboard of sightings and lifetime impact, device and sighting detail screens, and a community map with neighbour profiles and shareable summary cards.

Link to the interactive prototype.
©DHA 2026