Dynamic Machine Vision Solution for Improving Construction Site Productivity and Safety

Technological advancements are moving rapidly in the construction space, touching all areas of the ecosystem. With these new technologies and industry-wide digital transformation of paperwork, images, and video feeds, comes data – and lots of it.

This rich data creates a multitude of new opportunities for AI solutions – including Machine Vision and Intelligent Document Processing – to catalyze automation throughout the project lifecycle, from design management to preconstruction, prefabrication, construction, resource and equipment management, operations, scheduling and staffing, health and safety compliance, and project retrospectives.

Top construction firms today are eager to take advantage of these new technologies to accelerate innovation and drive competitive advantage in an industry where pennies – and seconds – matter.

Problem

Synaptiq’s client, a large multinational construction company, aimed to unearth new opportunities to improve efficiencies and reduce costs, and recognized the value of tapping into data within their own organization to do this. The client had high-resolution cameras on live construction sites which produced rich image and video data, and believed that by leveraging these assets in new ways they could improve cost and delivery efficiencies while also feeding into existing ERP systems.

Solution

Synaptiq used data from cameras (static images, video recordings, and live video feed) and built a machine vision model to monitor site productivity and audit site safety. The model was fed data on specific classes of construction objects and machinery, and then trained on that data, allowing the system to recognize and properly categorize and count instances of those objects on the construction site.

Importantly, the model was built to be flexible, so that the client could specific regions of interest and have the facility to add new types of objects in the future based on business needs. Our team provided documentation, training, and consulting on best practices for implementing and optimizing the model, which fed the data to a client-side dashboard. The client then used this dashboard to run reports, conduct analyses, and see patterns, so that they could make better-informed critical business decisions.

Impact

Our client has identified many immediate applications for this solution. For example, by unlocking the image and video data pixel-by-pixel, they could receive automatic alerts when a lumber shipment comes in, track who is on-site, and receive notifications if a contractor is not wearing PPE.

Because detailed logs for each object are available as outputs to be integrated into existing back-end systems for analysis, this solution facilitates real-time monitoring and alerts related to project management, scheduling, health and safety, security, and other purposes.

The client can use this solution to speed or automate processes based on the awareness the machine vision provides; it can help with scheduling, finance streamlining, notification of upcoming events, and other automations which can result in significant cost savings.

Interested in learning more or have a project idea? Please take a moment to fill out this form and we will be in touch shortly.

Additional Reading