Re-evaluating road defect classification and prioritization for AI systems - BC-996

Project type: Innovation
Desired discipline(s): Engineering - civil, Engineering, Engineering - other, Environmental sciences, Natural Sciences
Company: PavePal Technologies Inc.
Project Length: 4 to 6 months
Preferred start date: 03/01/2026
Language requirement: English
Location(s): Vancouver, BC, Canada
No. of positions: 1
Desired education level: Master'sPhDPostdoctoral fellowRecent graduate
Open to applicants registered at an institution outside of Canada: No

About the company: 

PavePal Technologies Inc. is a Canadian infrastructure-technology startup developing AI-powered tools for road condition assessment, asset management, and maintenance prioritization. The company leverages computer vision, geospatial analytics, and large-scale video data to help municipalities, road authorities, and infrastructure owners identify, classify, and prioritize road defects more efficiently and consistently.

PavePal’s platform processes geo-tagged imagery and video captured from vehicles and aerial systems to detect surface distresses such as cracks, potholes, and surface degradation. While the company has strong expertise in AI, data engineering, and software systems, it recognizes the critical importance of civil-engineering domain knowledge in ensuring that defect definitions, severity classifications, and prioritization logic align with accepted pavement-management practices.

Through this project, PavePal aims to strengthen the engineering rigor of its product by integrating pavement-engineering principles, standards, and decision frameworks into its AI-driven workflows, improving trust, interpretability, and real-world applicability for infrastructure decision-makers.

Describe the project.: 

The goal of this research project is to re-evaluate and strengthen the engineering assumptions used in AI-based road-defect detection and prioritization systems. PavePal currently applies machine-learning models to identify road surface defects from imagery; however, many of the underlying definitions, severity thresholds, and prioritization rules require deeper validation from a pavement-engineering perspective.

The intern will conduct a systematic review of pavement-management standards, manuals, and best practices (e.g., PCI-based methodologies, municipal and DOT guidelines) to assess how common road defects should be classified, grouped, and prioritized based on severity, risk, and intervention urgency. The research will focus on asphalt road networks in North America, with attention to how defect progression, traffic loading, and environmental factors influence maintenance decision-making.

Working closely with the PavePal engineering team, the intern will translate engineering knowledge into clear defect taxonomies, severity definitions, and prioritization frameworks that can be operationalized within data-driven systems. This includes identifying which defects are most critical to detect early, how multiple defects interact, and where AI-derived indicators may diverge from traditional engineering judgment.

The outcome of the project will be a validated defect-classification and prioritization framework, supported by references and engineering rationale, that can guide both AI model development and downstream maintenance planning. This work will improve the transparency, interpretability, and credibility of AI-assisted road-condition assessments.

Required expertise/skills: 

The ideal candidate has an academic background in civil engineering, pavement engineering, or infrastructure management, with experience or coursework related to pavement condition assessment, road maintenance, or asset management.

Required skills include:
● Knowledge of pavement distresses (cracking types, potholes, surface deformation, raveling, etc.)
● Familiarity with pavement-management systems and indices (e.g., PCI or related frameworks)
● Ability to interpret engineering manuals, standards, and municipal/DOT guidelines
● Strong analytical and technical writing skills

Assets (nice to have):
● Experience with infrastructure data, GIS, or asset inventories
● Exposure to data-driven or computational approaches in civil engineering
● Interest in applied research at the intersection of engineering and AI

Programming or machine-learning experience is not required; the focus of the role is on engineering judgment, validation, and translation of domain expertise into structured frameworks usable by software systems.