Fifteen people died on Calgary's streets in 2025. Six more died in the opening months of 2026. Behind those numbers are families shattered by routine trips to the grocery store or school. Traditional municipal planning usually reacts to a tragedy after it happens. City planners wait for the body count to pile up before installing a traffic light or a crosswalk flashers.
Calgary is trying a different approach. The City recently turned loose an artificial intelligence model to scan more than 17,000 uncontrolled intersections. Its goal is simple: find the danger zones before anyone else gets hurt.
The machine learning tool spat out a ranked list of the 50 highest-risk pedestrian intersections in the city. But having a list and fixing the problems are two entirely different battles.
Inside the Numbers of Calgary Pedestrian Safety AI
Most residents assume city planners know every dangerous street corner like the back of their hand. They don't. Traditional engineering relies on what the industry calls "conventional warrants"—strict formulas based on historical collision counts, traffic volumes, and pedestrian counts. If nobody has died at a specific corner yet, it's hard to justify spending money on it.
That framework is broken. It forces communities to wait for blood on the pavement.
The new machine learning platform flips that script. It assesses thousands of data points every six hours and updates its master model monthly. It looks at roadway geometry, lane widths, medians, weather conditions, time of day, proximity to schools and transit stops, and 2021 census figures. It assigns each uncontrolled intersection a risk score between 0 and 1000.
Out of 17,000 locations, the algorithm flagged 28 intersections with risk scores above 95, and 137 spots sitting above 90. Topping the list is the intersection of Martindale Boulevard and Martindale Gate N.E., closely followed by Falshire Drive and Falton Drive N.E.
When the Community Development Committee reviewed these findings, some local politicians scratched their heads. Ward 5 Councillor Raj Dhaliwal admitted he was surprised by the top-ranking spots in his area because he didn't personally see massive crowds of foot traffic there. But that is precisely why the algorithm matters. Human observation is biased by the times of day we happen to drive past a corner. Algorithms process raw, continuous variables without blinking.
The Budget Wall and Half-Measures
A brilliant algorithm is useless if city council won't fund the physical concrete and wiring required to fix the problem. That is where Calgary's initiative runs into a brick wall.
Out of the top 50 high-risk spots identified by the machine learning model, city administration only recommended safety upgrades for 26 locations. The estimated price tag for those specific fixes sits around $6 million.
Why stop at 26? Money.
Municipal budgets are tight, and councillors have to weigh pedestrian safety against every other line item. If the city wants to fix all 50 spots right now, or scale up to broader Vision Zero targets—such as addressing 150 to 300 locations annually under different funding envelopes—millions more must be squeezed from future budgets.
If approved during upcoming budget deliberations, design work for these high-risk areas could start in 2027, with physical construction stretching into 2028 and 2029. For someone who has already been hit by a vehicle in an unmarked crossing, waiting three years for concrete curbs and flashing beacons feels like an eternity.
Why Human Error Meets Algorithmic Sight
Technology can't solve everything on its own. Urban mobility experts point out that modern streets face a dual crisis of distraction. Drivers look at phones. Pedestrians look at phones. Commuters rush to and from work, cutting through dense residential zones where non-local traffic creates chaotic driving patterns.
Ward 8 Councillor Nathan Schmidt noted that high-density sectors deal with massive commuter pulses. The AI model captures that hidden friction by analyzing traffic flow dynamics, but politicians still need to balance the cold metrics with local resident feedback.
Relying entirely on software creates a trust gap. Residents want to know why their specific street corner missed the shortlist while a neighboring block got funded. Transparency is mandatory if the city expects taxpayers to support these multi-million-dollar safety overhauls.
What Needs to Happen Next
If you live in Calgary and want to see your local intersection prioritized, don't wait for an algorithm to flag it.
- Document near-misses and file 311 reports every single time a close call happens. The city's data intake relies heavily on public reporting to feed its risk-scoring inputs.
- Track upcoming municipal budget debates where the $6 million allocation for these 26 intersection upgrades will face council votes.
- Contact your local ward councillor to demand that safety interventions scale up to cover all 50 high-risk locations rather than stopping halfway.
Data-driven urban planning is only valuable when backed by political will and actual funding. Calgary has the map of where people are most likely to get hurt. Now it has to decide if protecting them is worth the price tag.
