Methodology · Edition 01 · April 2026

How PULSE works, and what it doesn't pretend to know.

PULSE is a neighbourhood-level Social Isolation Index for Edmonton, built as an operational layer on top of the City's existing Community Safety & Well-Being (CSWB) Strategy. This page explains every signal it fuses, every weight in the formula, and every known gap.

The Formula

PULSE's Social Isolation Index is a weighted composite of five normalized signals per neighbourhood. Each input is scaled to 0–100 before weighting, so the final SII sits on the same scale. Weights sum to 1.0. The current weighting reflects a v1.2 (April 2026) signal upgrade — see the changelog at the bottom of this page.

SII · per neighbourhood · 0–100
SII = 0.22 · Distress
+ 0.25 · AgeRisk
+ 0.25 · Silence
+ 0.13 · Mobility
+ 0.15 · LowIncome

# Alert threshold: SII ≥ 70
# Critical threshold: SII ≥ 85

Distress (22%)

311 needle-collection requests per neighbourhood over a rolling 30-day window, normalized against a ceiling of 30 incidents per neighbourhood per month. Needle reports are not a proxy for addiction or homelessness — they are a proxy for concentrated visible distress in public space. A neighbourhood where needles are being picked up is a neighbourhood where multiple social safety nets have already failed. Weight reduced from 0.30 to 0.22 in v1.2 to reflect the legitimate critique that needle-pickup is a substance-use proxy more than a loneliness proxy; the signal is retained because the two correlate at the neighbourhood level, but it no longer dominates the index.

Age Risk (25%)

Percentage of residents aged 65 and over, from 2016 Federal Census data (City Open Data dataset phd4-y42v), normalized against a ceiling of 25%. Social isolation risk scales disproportionately with seniors-living-alone rates, so elevated senior density is a structural baseline risk, not a direct measurement of loneliness.

Silence (25%)

The inverse of 311 service request volume. Counter-intuitively, low 311 call volume in a neighbourhood with known population is a warning sign, not an absence of problems. It typically means residents have stopped expecting a response, don't know the service exists (common in newcomer populations), or lack digital/English access. Baseline expectation: ~80 calls per neighbourhood per month.

Mobility (13%)

Inverted transit proximity score. Derived from ETS GTFS stop-density and neighbourhood centroid proximity. A senior without a car and without a bus stop within 400m is structurally more isolated than one with both.

Low-Income (15%)

Percentage of households earning under $40,000/year, sourced from the 2016 Federal Census neighbourhood-level open data (data.edmonton.ca, dataset jkjx-2hix). Normalized against a 50% ceiling — Edmonton's city-wide household-income distribution puts roughly 22% of households under $40k/yr; a neighbourhood at 50%+ is in concentrated poverty. Low-income status is a peer-reviewed isolation correlate (reduced ability to participate socially, smaller networks, fewer transport options) and is a structural factor that compounds every other signal in the index.

Crime — REMOVED FROM SII (v1.2)

Crime data from the EPS Community Safety Data Portal is still fetched and displayed in the neighbourhood panel for context, and remains a component of the Concentrated Distress Index (CDI) and a small component of the Senior Isolation Index (SII-S) for its safety-perception value. It has been removed from the main SII because EPS reporting is known to under-count incidents in newcomer communities — meaning crime data structurally under-weights exactly the neighbourhoods PULSE most needs to flag. Including it in the headline index introduced a bias the methodology could not honestly defend. Audit trail preserved here intentionally.

Composite Indices

Beyond the core SII, PULSE computes three composite sub-indices — each weighted combinations of the same five normalized signals, targeted at specific Edmonton agency partners. The purpose isn't to replace the SII, but to surface the particular risk profile of each neighbourhood in terms a given agency can act on.

Each index is a weighted 0–100 score. All weights are published here for audit.

Senior Isolation Index · SII-S

For: Edmonton Seniors Coordinating Council (ESCC), Alberta Health Services Seniors Health programs.

SII-S · 0–100
SII-S = 0.40 · AgeRisk
+ 0.30 · Silence
+ 0.20 · Mobility
+ 0.10 · Crime

Elevates the senior-specific isolation factors — a neighbourhood with high 65+ density, low civic engagement, and limited transit access is where ESCC's home-visit program has the highest prevention ROI. Crime is retained at a low weight because safety perceptions compound senior isolation (older residents stay indoors more in higher-crime areas).

Newcomer Vulnerability Index · NVI

For: Edmonton Immigrant Services Association (EISA), Catholic Social Services, Mennonite Centre for Newcomers.

NVI · 0–100
NVI = 0.40 · NewcomerProxy
+ 0.35 · Silence
+ 0.15 · Mobility
+ 0.10 · (100 − AgeRisk)

Puts the Invisible Cohort insight on a number. The inverse of AgeRisk is included because newer-settlement neighbourhoods typically skew younger — so a neighbourhood with low senior-density, high newcomer density, and high engagement-silence is the most likely to host spouses on dependent visas and newly-arrived families that civic data currently misses entirely.

Concentrated Distress Index · CDI

For: Boyle Street Community Services, EPS Community Safety Liaison, City Community Standards.

CDI · 0–100
CDI = 0.45 · Distress
+ 0.35 · Crime
+ 0.20 · Silence

Surfaces neighbourhoods where acute stressors compound — needle-reports, concentrated crime incidents, and residents who have stopped engaging with civic services because they no longer expect response. These are the neighbourhoods Boyle Street's outreach teams already deploy to; CDI provides a quantitative frame for prioritizing between them when resources are limited.

Hidden Loneliness Index · HLI

For: AHS Home Care, community leagues, friendly-visiting programs, LifeDojo mentorship.

HLI · 0–100 · v1.4
HLI = 0.35 · Silence
+ 0.25 · AgeRisk
+ 0.20 · Mobility
+ 0.15 · LowIncome
+ 0.05 · (100 − Distress)

The inverse of CDI. Every other index chases visible distress — the neighbourhoods where problems announce themselves. HLI targets the opposite profile: quiet, aging, transit-poor, low-income areas where nothing shows up in acute data because the people at risk never leave the house. The distress term is inverted — low street-level distress raises the score, because visible distress means services are already looking there. Quiet + old + stuck + poor, where nobody is looking, is where loneliness compounds until it becomes a wellness-check call. HLI is deliberately a prevention signal: its neighbourhoods should get friendly-visiting programs and social connection before anything measurable goes wrong.

Paired with HLI is the social-infrastructure layer (v1.4): a per-neighbourhood count of "third places" — community centres, libraries, places of worship, social facilities, cafés, and pubs — from OpenStreetMap. Neighbourhoods with HLI ≥ 55 and ≤ 1 third place are flagged as Event Deserts: structurally lonely areas with nowhere to gather. This is deliberately built from public infrastructure data, not personal or social-media data — PULSE measures the conditions for loneliness, never individuals.

All four indices use the same normalized signals as the main SII. No new data is introduced; the same transparent inputs are recombined with different weights to surface distinct risk profiles. Every weight is visible here. Every calculation is auditable.

The Invisible Cohort

PULSE's flagship signal. A neighbourhood is flagged as Invisible Cohort when newcomer-population density is high (newcomer proxy ≥ 60) AND Silence is high (≥ 60). The combination identifies populations the City's current 311-based civic data structurally cannot see: immigrant spouses on dependent visas, first-generation elders, residents without digital or English access to civic services.

The WHO Commission on Social Connection (2024) and the Harvard Study of Adult Development (75 years, 724 lives) together establish loneliness as a mortality factor comparable to smoking fifteen cigarettes a day. The Invisible Cohort is where that mortality risk concentrates most in Edmonton — and where civic intervention is least likely to reach without a targeted signal.

Data Sources

SignalSourceDatasetLive?
311 Requests City of Edmonton Open Data q7ua-agfg Yes · Socrata SODA API
Needle Collection City of Edmonton Open Data 6k27-tnmh Yes · Socrata SODA API
Age Distribution Statistics Canada via City of Edmonton phd4-y42v (Census 2016) Yes · queried on boot
Neighbourhood Boundaries City of Edmonton Open Data 65fr-66s6 (GeoJSON) Yes · Socrata
Crime Incidents Edmonton Police Service Crime_Incidents FeatureServer Yes · ArcGIS REST API
Transit Proximity ETS GTFS + neighbourhood centroid derived Partial · estimated baseline, live data pending ETS MoU

Known Gaps

PULSE is only as honest as its sources. Several high-value signals are not yet in the index because Edmonton publishes them at city-wide level, behind FOIP, or in non-queryable formats. These are not weaknesses to hide — they are the roadmap for what a pilot partnership could unlock:

CSWB Alignment

Edmonton's Community Safety & Well-Being Strategy tracks seven pillars at the city level. PULSE maps each of its signals to the pillar it operationalizes at neighbourhood granularity:

What PULSE Is Not

PULSE is not a clinical tool. It does not diagnose loneliness. It surfaces neighbourhood-level risk patterns from already-public data.

PULSE is not a replacement for CSWB. CSWB is the Council-endorsed strategy. PULSE is an operational layer that makes the same signals actionable at neighbourhood granularity.

PULSE is not a black box. Every input, weight, and threshold is published on this page. Reviewers can audit the methodology end-to-end.

PULSE is not finished. This is Edition 01. Edition 02 will ship when a pilot partnership unlocks one or more of the gaps above.

About

PULSE is a project of Dojo Foundations, an Edmonton-based civic-tech studio founded by Sai Praneeth Chandra Balla (Dojo). Dojo's background is in industrial SCADA — the monitoring systems that have run power grids and oil refineries for four decades. PULSE applies the same architecture to municipal wellbeing data. Sister project: LifeDojo, a mentorship platform that acts on the isolation PULSE measures.

The project exists because Edmonton's data already contains the answers its agencies need — they just aren't queryable at the altitude where service happens. Someone had to build the operator view. That's PULSE.

Contact: [email protected] · dojofoundations.com · Live dashboard