Smart Inspections Digital Assistant (InspectorAI)
Modernizing building inspections with AI to boost efficiency and compliance.
Project Overview
Updated November 5, 2024.
The Problem
British Columbia is projected to need 600,000 new homes by 2030. To achieve this goal, innovative new tools must be adopted to boost construction productivity while maintaining high-quality standards and complying with Building Code requirements. Building inspections are crucial checkpoints in new housing construction, where individuals verify compliance with the proposed design and all applicable Building Code regulations and standards. Deficiencies must be corrected immediately to avoid the risk of harm to building occupants or nearby property and residents.
Building Inspectors use their knowledge and experience accumulated over the span of their careers to identify deficiencies in construction. The most experienced inspectors can often predict where issues may occur before they happen and focus attention on those specific areas. It’s a form of risk management that has been employed regularly over time. However, with 50% of Canada’s building officials planning to retire the profession in the coming years, this key pool of experts is being severely depleted. An AI solution trained on the experience of hundreds of building inspectors across Canada will equip newer officials with the tools necessary to maintain the highest safety standards.
How We Are Solving It
InspectorAI will use advanced artificial intelligence (AI) to learn from past inspection data and provide site-customized predictions of building code issues. The real-time information gathered will be used to improve construction compliance and reduce work stoppages, allocating resources where they are most needed. InspectorAI will also automate routine tasks like recording information and communicating with builders, freeing inspectors to focus on their core duties and inspect more homes under construction. Led by Cloudpermit, in collaboration with TedfordAI and Alberta Machine Intelligence Institute (Amii), this project aims to modernize the way building inspections are performed and remove the delays that can result from using current methods. Most jurisdictions still employ paper-based approaches to scheduling, documenting, and tracking inspections at various stages of construction. This takes up valuable time for preparing visits, writing inspection notes, notifying builders of statuses, and ensuring deficiencies are resolved at future visits. It’s easy for issues to “fall through the cracks” and cause future consequences for homeowners.
InspectorAI will use Natural Language Processing (NLP) to convert speech to text for recording notes but then link those notes to specific Building Code requirements so that they can be accurately documented and shared with the builder.
Municipalities also face challenges in quickly and efficiently overseeing the construction of new housing, related administrative tasks, and the associated volume of data. For the few cities that have started exploring the use of AI for compliance inspections, it has taken them up to 10 years to collect the needed quantity of data. The project’s data platform, Cloudpermit, can collect the same volume in a matter of months, allowing for the training and testing of ML models to accurately predict which deficiencies will be found on specific sites and at which stage during the construction of new housing. This will enable local governments to better utilize their staffing resources and ensure safe and high-quality homes are constructed in communities.
The project will also develop a builder application that enables scheduling inspections and receiving notifications of results. It will use the predictive ML model to direct builders towards any areas where deficiencies are likely to occur, allowing them to address these issues before an inspector arrives. This helps avoid potential work stoppages and boosts productivity for both inspectors and municipalities. By overseeing and automating many tasks involved in preparing, conducting, and reporting on inspections, it will help reduce stoppages caused by non-compliant construction, ensuring homes are built on time.
British Columbia is becoming a North American leader in digital construction and electronic permitting. Adding a first-of-its-kind product like InspectorAI into the industry takes that work to the next level, keeping B.C. and Canada at the forefront of innovation and AI-driven success.
