03, Spatial data & platform
OrganisationThe Crown Estate
Marine Data Exchange
Connecting marine investigation with data contribution
A large marine, offshore and environmental repository is only useful when people can find their way into it. The work connects search and spatial investigation with the quieter task of preparing data for review.
Product design · Marine data platform · Figma design and prototype
Project brief
About
Marine Data Exchange is The Crown Estate's platform for offshore survey data and research. The project connects the discovery and interpretation of marine data with the work of preparing submissions for review.
Problem
The challenge was to help people locate relevant marine evidence without already knowing the exact dataset or search term, while giving contributors a clear route through metadata, file validation and review.
Objective
Connect search, geographic context and dataset detail, and make submission requirements and recovery steps explicit.
How the interfaces support the task
Desktop discovery and contribution: Search and explore marine evidence through geographic context and dataset detail, or prepare a submission through metadata, file validation and review.
Tablet and mobile layouts: Keep dataset context and status readable as filters collapse; a focused location-details sheet brings inspection above the map. Contribution layouts prioritise the current step, requirements and recovery action.
Together: These are responsive views of the same service, with reading and contributing treated as distinct tasks.
Scope, work needed
- Search and spatial discovery across wind, tidal, wave and species domains
- Map and inspector composition with time range, location detail and supporting charts
- Dataset detail and context hierarchy for readers with different levels of expertise
- Contribution journey: details, metadata, files, validation, review
- File-validation pattern including a blocking coordinate-reference error and its recovery
- Shared interface patterns and responsive layouts for discovery, submission and QA
Personas
Research user
TaskFind relevant marine evidence and preview a dataset before committing to a download.
Design responseSearch, filters and the inspector preview the record, coverage, species, format, before any download step.
Contributor
TaskGet device access, understand what a submission needs and see where it stands.
Design responseSubmission guidance sits with the form, and status is a first-class state rather than an email thread.
Quality assurance
TaskResolve incomplete submissions and record what validation found.
Design responseValidation and resolved states show the specific gap, the action taken and who is waiting on what.
Operations
TaskPrioritise the queue and keep oversight of what is ready to publish.
Design responseQueue and publication views order work by state and age, keeping the next decision visible.
Design personas · Portrait · Troy Spoelma / Unsplash; Portrait · The Connected Narrative / Unsplash; Portrait · Christina @ wocintechchat.com / Unsplash; Portrait · Dorrell Tibbs / Unsplash.
Our contribution
Search and spatial discovery design, so a reader can arrive without a dataset name and still narrow towards relevant marine evidence.
Dataset and context hierarchy across the intelligence domains, keeping location, coverage, recency and review status beside the record.
The contribution and validation flow, from project details and metadata through file checks to review, with blockers named and given a recovery action.
Review and QA states, shared interface patterns and responsive layouts for both the reading and the contributing tasks.
Design narrative
01Let people find evidence without knowing its name
Search, filters and geographic context work together, so someone who knows the region or the domain but not the dataset can still narrow towards relevant material. Results keep location, recency and review status beside each record.
The desktop tidal view then brings map controls, a location inspector, time context and supporting charts together. The map gives orientation, the inspector makes the selected location and its source explicit.
Tidal intelligence, map orientation with an explicit location inspector and source model.
Species intelligence, the same map-plus-inspector model applied to another domain.
02Change the composition when the task changes
Submission uses the same product language with a calmer hierarchy. Files, validation messages and progression sit together, so a contributor can see what is required and what is outstanding in one place.
A missing coordinate reference system is named as the blocker, with a resolution action beside it and the route to review disabled until it is cleared. The block is explained rather than simply enforced. Once resolved, review opens and the submission timeline records how the record got there.
Project details stay visible through submission and review, so a returning contributor can check the identity, survey context and date range before continuing.
Validation, the blocking issue names its cause and sits beside its resolution.
Files, sizes and check results are demonstration data. There is no upload or validation service behind the screen.
Resolved, review only becomes available once the blocking issue is cleared.
Survey identity, dates and coverage are illustrative scenario continuity, not historic records.
03Carry the same meaning into a smaller view
The desktop contextual inspector becomes a focused location-details sheet on mobile, sitting above the map rather than shrinking the whole desktop interface. Search behaves the same way: filters collapse, but a dataset keeps its location, recency and review status.
Contribution gets its own task hierarchy at phone width, with the current step, its requirements and the next action kept in view.
Mobile, the inspector becomes a sheet above the map rather than a shrunken desktop.
Search on mobile, filters collapse, but status and dataset context stay.
Search on tablet, results keep their review status beside the dataset.
04Make repetition useful
Map inspection, file validation and review patterns connect the component library to recognisable tasks, so the same behaviour is reused wherever those tasks appear.
The architecture page holds it together: navigation, the shared information model, the submission state model and the journeys that cross them.
Patterns are documented against the task they serve, not as abstract swatches.
Library inventory. Not an adoption metric.
One shell, intelligence, contribution, QA and stewardship written as one architecture.
Product architecture and journeys as designed.
05What was delivered
Discovery, spatial and dataset compositions across the intelligence domains, and a full contribution path from details through validation to review.
A shared pattern library, responsive layouts for both tasks, and an interactive prototype covering the intelligence, submission and QA sequences.
Systems & responsive
Design system
Reference, colour and scale variables support the shared patterns: map inspection, file validation and review status. Each pattern is documented against the task it serves, so a validation state or a status label reads the same way in discovery, submission and QA.
Responsive behaviour
The desktop contextual inspector becomes a focused location-details sheet on mobile, keeping the map and primary navigation. Search collapses its filters while keeping dataset status visible, and contribution takes its own task hierarchy at phone width rather than reusing the reading layout.
Prototype
Prototype walkthrough
Marine data workspace walkthrough
Desktop discovery, review and contribution sequence.
1:40
Explore wind, coastal sensor and tidal views, then move through QA review, data submission and operations.
Recorded prototype walkthroughs with illustrative data.
The prototype covers the discovery, submission and QA sequences, including the blocking validation state and its resolution. It is an interactive design, so there is no upload, geospatial engine or review service behind it.
Explore prototype routes
Gallery
25 images
1 / 25
Overview: every marine domain, recent submissions and system status on one entry screen.
Assets, datasets and submission records are demonstration data.
Project notes
A short set of notes on the material shown in the screens.
- Environmental readings, datasets and submission records shown are demonstration data.
- Survey identities, regions and dates in the submission example are illustrative prototype content.
Takeaway & evaluation
A large repository becomes approachable when search, geography and dataset detail form one path, and when a blocked submission names its cause and the step that resolves it.
Outcome metrics not available.
What to measure
Proposed evaluation, not conducted research.
- Can someone with limited domain knowledge locate a relevant dataset through search or the map, without already knowing the exact dataset name?
- Can a contributor resolve a blocking validation state, such as a missing coordinate reference system, from the recovery action alone?