An AI-Enabled Condition Assessment of the Lake Winona Bike Path

Presenter(s)

Anton Kadlec

Abstract

This project was developed as part of the Winona State University Artificial Intelligence Innovation and Engagement Pilot Fund, which supports extracurricular AI projects that address campus and community challenges and strengthen career readiness through hands on experience. This project is presented as part of the Warrior.AI Symposium.

This project develops an AI-powered trail condition assessment for the Lake Winona Bike Path, a 5.3-mile paved loop encircling the East and West Lakes in Winona. The system correlates three data sources to produce cost efficient maintenance recommendations for city staff and improve Return on Investment: International Roughness Index (IRI) scores from Total Pave, Pavement Condition Index (PCI) scores collected through a field survey, and community experience feedback gathered via QR Code surveys posted at five trailhead locations. IRI measures pavement roughness in inches per mile using a mobile phone sensor and GPS. Higher IRI values indicate rougher surfaces with scores generally ranging from 0 to 500. PCI rates overall pavement distress on a scale from 0-100 where 0-10 is a failed surface, 10-25 is a very poor surface, 25-40 is a poor surface, 40-55 is a fair surface, 55-70 is a good surface, 70-85 is a very good surface, and 85-100 is an excellent surface. Together these two quantitative indicators provide a comprehensive picture of trail health at the segment level.

The project draws on my three years of professional infrastructure experience as an intern at GoodPointe Technology, a St. Paul based infrastructure asset management firm. This work included pavement condition data collection and management for municipalities including Apple Valley, Eden Prairie, and the St. Cloud Area Planning Organization, where I collected 150 miles of IRI data during Summer 2024.

The AI advisory system is built using Microsoft Copilot Studio. Survey responses are correlated with IRI and PCI measurements to produce a serviceability score for each trail segment. This allows the city staff to identify where their limited maintenance budgets will have the greatest impact.

College

College of Business

Department

Business Administration

Campus

Winona

First Advisor/Mentor

Pat Paulson

Location

Kryzsko Great River Ballroom, Winona, Minnesota; United States

Start Date

4-23-2026 1:00 PM

End Date

4-23-2026 2:00 PM

Presentation Type

Poster Session

Format of Presentation or Performance

In-Person

Session

2a=1pm-2pm

Poster Number

27

Share

COinS
 
Apr 23rd, 1:00 PM Apr 23rd, 2:00 PM

An AI-Enabled Condition Assessment of the Lake Winona Bike Path

Kryzsko Great River Ballroom, Winona, Minnesota; United States

This project was developed as part of the Winona State University Artificial Intelligence Innovation and Engagement Pilot Fund, which supports extracurricular AI projects that address campus and community challenges and strengthen career readiness through hands on experience. This project is presented as part of the Warrior.AI Symposium.

This project develops an AI-powered trail condition assessment for the Lake Winona Bike Path, a 5.3-mile paved loop encircling the East and West Lakes in Winona. The system correlates three data sources to produce cost efficient maintenance recommendations for city staff and improve Return on Investment: International Roughness Index (IRI) scores from Total Pave, Pavement Condition Index (PCI) scores collected through a field survey, and community experience feedback gathered via QR Code surveys posted at five trailhead locations. IRI measures pavement roughness in inches per mile using a mobile phone sensor and GPS. Higher IRI values indicate rougher surfaces with scores generally ranging from 0 to 500. PCI rates overall pavement distress on a scale from 0-100 where 0-10 is a failed surface, 10-25 is a very poor surface, 25-40 is a poor surface, 40-55 is a fair surface, 55-70 is a good surface, 70-85 is a very good surface, and 85-100 is an excellent surface. Together these two quantitative indicators provide a comprehensive picture of trail health at the segment level.

The project draws on my three years of professional infrastructure experience as an intern at GoodPointe Technology, a St. Paul based infrastructure asset management firm. This work included pavement condition data collection and management for municipalities including Apple Valley, Eden Prairie, and the St. Cloud Area Planning Organization, where I collected 150 miles of IRI data during Summer 2024.

The AI advisory system is built using Microsoft Copilot Studio. Survey responses are correlated with IRI and PCI measurements to produce a serviceability score for each trail segment. This allows the city staff to identify where their limited maintenance budgets will have the greatest impact.