One front door for Sweden's OOH measurement stack

This hub connects the two pillars of OOH measurement in Sweden: the measured trips of real human respondents that powers live mobility dashboards, and the new Agent-Based Model (ABM) that uses a synthetic population to model behavior at DeSO resolution. Use the respondents' trips for operational insights today, and ABM for the next-generation, behavior-rich measurement foundation.

Models

Choose your entry point

Data inventory, respondent models, population synthesis, their behaviour, everything under one umbrella for integrated OOH planning and evaluation.

Real data

Inputs and sources

Tens of data sources collected from Swedish national institutions (SCB, Trafikverket, ...). Measured trips of thousands of respondents. The live dashboard suite with mobility playback, transit observatory, and OOH impression analytics derived from respondent travel behavior.

Complete metadata catalogue
OOH impressions and mobility insights
Transit flow and network analytics

Enter

Synthetic data

Population synthesis

A synthetic, full-population human layer that powers behaviour-based exposure measurement and advanced audience segmentation.

DeSO-level population synthesis
Human Layer and ABM report surfaces
QA/QC reporting

Enter

Synthetic data

Agent-Based Model

Agent-based model expands the respondents' routes identifying reason to travel to the realistic synthetic population. adding most probable routes at most probable time with reactions to external shocks like weather events.

Realistic trips coming from hierarchical plan of the week, day in the week and intraday plan
Able to accomodate shocks and disruptions

Enter

Why all those models

Research, short-term insights, long-term foundation.

Abundance of demographic and socioeconomic data together with real human trips measurements enable projecting the real data into a durable, data-rich foundation to model behavior across time, demography, and geography. Together the models deliver both immediacy and depth.

Operational speed OOH impression dashboards deliver immediate OOH planning and performance insights.

Unprecedented realism Agent-based model stems from latest advancements in population modelling and sheer data availability.

Behavioral depth Activity-based model captures the human context behind exposure, beyond traffic counts.

Strategic continuity Both systems align to the same OOH measurement language and outputs.

Roadmap

Converging toward unified measurement.

The ABM roadmap shifts us from a respondent-weighted view of mobility towards a full-population simulation with stronger validation and much richer audience segmentation, while the respondents trips remains the workhorse for live operational delivery. We begun with a disciplined data inventory, bringing together national statistics on population and economic activity alongside traffic intensity models and related mobility layers. That foundation lets us move from “what we measured” to “what plausibly happens everywhere”, with clear traceability back to observed evidence.

Using previously measured trips from around 6,000 people, we built an OOH impressions calculator that does more than aggregate counts. It takes individuals and their journeys and expands them into full visualisations where agents traverse OOH visibility zones, allowing impressions to be estimated in a spatially and temporally realistic way. Critically, those measured trips are not treated merely as weighting factors. They are behavioural anchors: signals about routines, constraints, preferences, and context that we preserve and generalise rather than smoothing away.

On top of this, we constructed a highly realistic synthetic population, locating homes, workplaces, schools, and—where relevant—holiday homes, and anchoring those agents across Sweden with plausible geographies and household structures. We then used the measured trips to build a hierarchy of planning: first a “type of week” driven by the person and household profile, with shocks such as holidays and festivities; then archetypal day-of-week patterns tailored to the individual; and finally an intraday plan. This is where the model becomes recognisably human: the early start to escort children before work, the evening trip to the pub, the habitual cycle commute, or the Thursday departure to a lakeside holiday home to telework in quieter surroundings.

These behaviours are not hand-waving. They are grounded in extensive published research on human activity patterns and in our own inference of trip purpose from the measured data, including semantic segmentation of trips to identify likely reasons for travel. The result is an exceptionally robust population model whose uses extend well beyond OOH measurement: it provides a general-purpose behavioural substrate for policy testing, scenario analysis, and any application where realistic activity, movement, and exposure need to be simulated at population scale.

Human Layer ABM reporting Live dashboards OOH impressions

Project status

Accomplished steps:

Data and research articles collection

Visualisations of input data

Population synthesis incuding workplace, school, weekend houses

Population validation and QA/QC reporting

Behavioural archetypes for every person and household, every week type, region and urbanisation level types

Next steps:

Trips synthesis for every person

Trips fitting into a realistic national ABM model, validation

Introduction of shocks and disruptions (weather, strike, ...) into the model

Fully operational ABM for various scenario modeling