Frequently Asked Questions

The product centrally collates information about parks, recreation, and sport facilities across Canada, including location, size, amenities and age. This information drives additional insight provided by built-in:

• Dashboard highlighting infrastructure and facility metrics and benchmarks, including flexible and per capita counts at the national, provincial/territorial and local levels.
• Movement mapping products illustrating the casual, typically non-measured, use of active environments over time and across periods of the year.
• Community profile mapping illustrating populations, health, and wellness demographic profiles at neighbourhood, city, and provincial/territorial geographies across the country.
• Ability to make infrastructure comparisons across provinces, territories, municipalities, and neighbourhoods.
• Determine the proximity of facility provisions to various equity-deserving populations and demographic groups.

The product has been designed specifically for individuals and organizations involved in the sport, physical activity, recreation, and parks sector. Groups that may find value within the product include:

• Municipalities
• Facility operators
• National, provincial/territorial, and community sport organizations
• Not-for-profits serving the sport and recreation sector

If you or your organization are involved in the sport, physical activity, recreation, and parks sector, the product has been built for you. This product can be of use in other sectors, including wildlife and conservation, emergency response and preparedness, and others. Reach out to intelligence@activexchange.ca to learn how this product can be used.

With thousands of sport and recreation sites and facilities across Canada, it is common to find a specific site with missing, incomplete, or incorrect data in the platform. Therefore, we encourage you to add and update the site and facility information owned or operated by your organization and for any site or facility you frequently visit.

As an open-sourced dataset developed by the community, for the community, the accuracy of the dataset is ensured primarily by the users of the product. ActiveXchange and CPRA play important roles in facilitating effective and accurate upload and usage of the product and monitoring for any large-scale inaccuracies contained. The seed dataset from Statistics Canada, Open Data Recreation Sport Facilities, underwent rigorous verification with other public sources prior to starting the product.

The CAN Active Spaces product contains many different infrastructure types relevant to the sport, physical activity, recreation, and parks sector. Example site and facility types include:

• Recreation Centres
• Gyms
• Sports Courts
• Ice Rinks
• Community Halls
• Trails (Hiking & other)
• Parks
• Fields & Pitches
• Green Spaces
• Golf Courses
• Other spaces supporting the sport, physical activity, recreation, and parks sector

If you’re unsure if your site or facility should be included, please contact ActiveXchange at intelligence@activexchange.ca.

Facilities are included when municipalities or organizations add or upload their infrastructure data to the platform. Anyone can create a free account and data. Support is available to help with formatting and data uploads of large datasets.

Movement is a privacy-forward dataset of device density, activity, and movement over time drawn from 700M+ users globally contributing 20B+ live location updates daily, drawn from 45,000+ smartphone apps. These apps provide ‘event pings’ on an ongoing basis, which is the sharing of GPS locations (even if the app isn’t open on the screen, depending on app/phone settings). GPS data also means devices can provide location data where there is no cellular coverage. Due to the number and breadth of these apps, the data becomes representative of the whole population and removes the influence of different demographics or different geographic areas (84%+ accuracy when validated against known counts and economic activity).

Hi-resolution location accuracy sourced from on-device GPS. This also means the data is less affected by areas with no cellular signal.

An underlying panel of millions of monthly active users in every major market across multiple travel modalities (legacy data sets are limited to relatively small sample sizes or based on fixed sensors with limited insights and no coverage outside urban areas).

Sourced from a diverse mix of thousands (45,000+) of different apps (other data sets are tied to a small number of sources which causes bias in the data).

The most comprehensive privacy-forward global dataset, complying with Canadian privacy laws.

Movement data is captured autonomously using location services from mobile devices (for example, GPS-based signals). No physical infrastructure, sensors, or hardware installations are required. The data is aggregated, anonymized, and privacy compliant.

The dataset captures significant GPS-enabled device activity aggregated into geographic tiles of 100-metre resolution (quadkeys) that cover the entire region. Activity within these quadkeys or across collections of quadkeys can then be tracked on a rolling basis, with data provided back to January 2022 to allow for trends.

The data is all anonymized to be privacy compliant and normalized to create a linear index. The index is normalized to a baseline, where the index value 1.0 represents the 99.9th percentile activity level in a day/tile of January 2021 in each country. So, an activity index of 0.1 indicates the tile is 10% as busy (level of activity/visits) for this location in contrast to the baseline activity.

This does not include dwell time data or data on individuals, but it does provide nationwide coverage across all types of infrastructure and environments, from which significant relative insights can be drawn. For an individual to generate an ‘event’ (app shares GPS) which is linked to a quadkey varies by mobile app and there is no minimum dwell time. Across all apps aggregated, this provides a robust overall measurement.

Activity Trends provide a large-scale view of where and when people move through a geography. In particular, this data allows you to:

• Understand the relative changes in mobile device activity in space and across time.
• Visualize changes in device density by time of day, weekdays/weekends, across multiple months or years.
• Observe significant anomaly events that occur and correlate changes in activity before and after – this provides an alert of the potential need for action e.g. movement suddenly drops at a facility/space/path.

The Activity Index is the density metric which reflects the level of activity in the time window and spatial regions defined above. A normalization process is applied to smooth out any effects of extreme outlier activity and to reflect the general population (not just app and smart phone users). ActiveXchange calibrates public and known ground-truth data to equate the index to a useful estimation of real activity (i.e. people & time), accommodating different settings.

Good calibrations are always difficult and that’s especially true here: actual ground-truth measurements of that density-of-humans number simply do not exist at the very fine spatial and temporal scales required. Other similar data sets exist from other providers, and we compare Mapbox (ActiveXchange) data to the Economist’s Normalcy Index. Importantly, ActiveXchange undertakes further correlation validation with known local community sport and leisure users and attendance patterns in different settings and communities to ensure confidence of use.

Activity Trends data has a significant advantage over our competitors in the “calibrating device density” game: Mapbox-powered maps are used by a great number of different apps, each targeting a different subset of the population, running on both major mobile device platforms (iOS and Android). Localized adjustments (urban vs rural) are validated through use of an auto-scaling adaptive model.

ActiveXchange undertakes correlation and validation checks with a range of registered sport and leisure known activity. Using this, ActiveXchange will also be supporting clients to equate the index to approximate visit numbers.

Movement data does not directly identify individual devices, nor can demographics be determined for privacy and feasibility reasons. Instead, the GPS signals sent by devices through the associated applications requiring location services is collated to derive the activity index.

There is no way to determine the representation of any demographic from the Activity Trends source data.

Mobile device ownership is very high across age groups in Canada, and movement data is intended to complement, not replace, other data sources such as surveys or manual counts.

The anonymous telemetry data that provides the foundation for Movement Data is closely correlated with the movement of people, however, does not provide a direct measure of the absolute number of people or their movement patterns. As a result, the data must be presented in a unit-less activity index. Decisions made from Movement data should be informed by a comparison or correlation, whether that’s a comparison of the activity difference between two facilities/spaces on any given day, the change in activity for a given location over time, or correlation with other sources such as trail counters, and any combination of these. As ActiveXchange receives known registered or ground-truth data an approximate guide between likely users and activity index can be provided.

For any single place and time, the actual value of the activity index is not the only thing that matters. We also care about the difference in activity at that place and time relative to some other place or time. In fact, it makes more sense to talk about our confidence level for a comparison made from Movement data over space or time than of an individual data point.

ActiveXchange undertakes correlation and validation checks with a range of registered sport and leisure known activity. Using this, ActiveXchange will also be supporting clients to equate the index to approximate visit numbers.

Data is normalized to the base of the 99.9^th percentile activity index level of January 2021.  This ensures that an activity index of 1.0 represents the same density of GPS signals in 2021 as it does today.  Other normalizations practices include using baseline GPS signal frequency and similar parameters, which are conditional on the specific applications the signal is derived from. Outlier data and GPS signals located on private property (based on publicly available property designation sources) are cleaned from the data, to ensure the activity index is not biased by static devices (e.g. not in use, charging, etc.) over significant periods of time, while GPS signals may remain active.

Mobile device adoption and usage patterns vary meaningfully across urban, suburban, and rural residents. In addition, many mobile apps are designed for use exclusively in urban areas (e.g. ridesharing, food delivery, community/neighbourhood activism apps). This generates a significant imbalance in mobile device activity levels measured in urban vs. rural areas, artificially inflating urban activity levels and depressing rural activity patterns. To address this inconsistency, an auto-scaling adaptive model is applied that identifies and measures the urban/rural imbalance and applies a set of “corrections” at the city level on a daily basis. This self-adapting system is fully automated.

This dataset captures direct movement and GPS-signals from unique devices/apps where informed consent has been granted, and thus privacy compliant. The data is additionally anonymized at source for additional privacy compliance.

The data of individual devices is cleaned, de-duplicated (one device, multiple apps), and made available in a “big data” environment, allowing for highly detailed resolution and aggregation.

This enables the reporting on exact device counts, behavioural metrics like dwell time, speed and direction of travel, entry/exit points, pathing, etc. From these insights, point of origin, associated visits (before/after) and cross visits can be modelled as well.

While rates of informed consent represent only a proportion of total devices in a market, this proportion greatly exceeds sampling rates of more traditional surveying techniques.

Visit Reports provides a high-definition view of how, where and when people move through a specific geography. In particular, this data allows you to:

• Understand visit and visitor counts over time, by date, day of week, portion of day, and dwell time.
• See where visitors come from, the route(s) taken to arrive and depart, and their use patterns within a specific geography.
• Point of Origin data can then be used to correlate visitor metrics to demographic and other inferred characteristic information on population segment(s) served.