Background
Mott MacDonald, a global engineering consultancy, was tasked with developing a master plan for a complex transit-oriented development (TOD) in Manila, Philippines. The 12-hectare site sits at the intersection of two proposed railway alignments: the Metro Manila Subway Project (MMSP) and the North South Commuter Railway (NSCR)—creating a unique opportunity for integrated urban development.
David Turner, Transport Modeling Lead for Mott MacDonald in the Czech Republic, led the transport planning component. The project aimed to create a true "15-minute city" concept with skyscrapers, integrated cycling infrastructure, bus interchanges, and a carefully designed road hierarchy—all while prioritizing sustainable transport modes.
Given Manila's notorious traffic congestion—where one-hour traffic jams are routine—the sustainability focus was critical to the project's success.
The Challenge
The project required extensive transport data for building micro-simulation models and understanding traffic patterns across different times and days. However, traditional manual surveying methods presented significant obstacles:
- Time-intensive process: Manual video review would require approximately one month—two weeks for a junior team member to count footage, plus another week to convert data into usable formats
- Accuracy concerns: Even with extensive manual effort, the results would be inconsistent and require significant data cleaning
- Complex vehicle classification: Manila's diverse transport ecosystem includes unique vehicles like jeepneys (converted American army vehicles used for public transport), electric tricycles, and varied two-wheeled vehicles
- Limited pedestrian data: Traditional surveys typically ignore pedestrian trajectories, missing critical data for TOD planning
- Tight project deadlines: The month-long manual process simply wasn't feasible within the project timeline
"It would take a couple of weeks of a junior member going through camera footage definitely, and then it would take another week probably of converting that into a format that we can use... I think we're talking a month of work versus two weeks of work."
— David Turner, Transport Modeling Lead, Mott MacDonald
The Solution
Mott MacDonald chose GoodVision's AI-powered traffic analytics platform—a decision informed by David Turner's several years of familiarity with the technology and, critically, GoodVision's proven track record in the Philippines with local partners.
Key Capabilities Deployed
- Rapid automated processing: Video analysis was "instantaneous" compared to manual methods, providing turning movement counts and traffic flow data in a fraction of the time
- Custom Philippine vehicle classification: Pre-trained AI models included specific custom vehicle classes for Filipino transport modes, including jeepneys, eliminating the need for additional machine learning setup
- Automatic mode detection: The platform automatically classified vehicles (cars, two-axle vehicles, bikes, jeepneys, tricycles) without manual intervention — see the full vehicle classification guide
- Pedestrian trajectory analysis: GoodVision captured pedestrian movement data—"something that normally we ignore" in traditional surveys and providing crucial insights for TOD planning
- Visual reality features: The platform's visualization capabilities proved invaluable for stakeholder engagement, demonstrating Manila's chaotic traffic conditions with concrete evidence
- Exceptional technical support: The GoodVision team worked closely with Mott MacDonald team to define trajectories, clean data, and address implementation challenges earning a "10 out of 10" rating for responsiveness and attention
Results & Impact
GoodVision's AI platform delivered transformative results for the Manila TOD project:
|
Metric |
Result |
|
Time Reduction |
Approximately 50% (from ~1 month to ~2 weeks) |
|
Data Quality |
Significantly improved accuracy vs. manual counting |
|
Processing Speed |
"Instantaneous" video-to-data conversion |
|
Additional Insights |
Pedestrian trajectories (not captured by traditional methods) |
|
Client Satisfaction |
10/10 support rating |
Beyond Time Savings: The Accuracy Advantage
While the 50% time reduction was substantial, David Turner emphasized that improved accuracy was the most valuable benefit:
"More importantly, it's kind of the accuracy, I think. Even if we spend all that time manually watching videos, it does not yield good results now in our findings."
The combination of speed and accuracy enabled Mott MacDonald to "get to the answers that we need quite quickly"—critical for tight project timelines and client deliverables.
Conclusion
GoodVision's AI-powered platform proved to be the ideal solution for Mott MacDonald's complex Manila TOD project. The technology's pre-existing Philippine market knowledge, combined with rapid processing capabilities and exceptional support, enabled the consultancy to deliver high-quality transport analysis in half the time required by traditional methods—while achieving superior accuracy.
For transportation planners and engineers working on complex urban development projects, particularly in markets with diverse vehicle types and challenging traffic conditions, GoodVision offers a transformative approach to traffic data collection and analysis.
About GoodVision
GoodVision is an AI-powered traffic analytics platform serving 200+ transportation authorities globally across four continents. Our technology transforms video footage into actionable traffic insights, helping cities and consultancies make data-driven decisions about urban mobility and infrastructure development.