Skip to content

A First Look at the Gilbane-Nextera Robotics Platform: “Didge”

September 22, 2021
Eric Cushman
As part of its innovation strategy and commitment to advancing construction management, Gilbane with its joint venture partner Nextera Robotics, has developed a transformational artificial intelligence platform utilizing a fleet of autonomous mobile robots. These robots automate construction management, allowing our teams to track project progress and provide exceptional safety monitoring at job sites.

New AI Platform Provides Cost-Efficient Construction Management Tool

The new robotics platform, dubbed “Didge”, provides autonomous “eyes-on-site” by obtaining detailed visual data at construction sites, creating a bridge to the field between design models, digital twins and plans. The name Didge is derived from the platform’s capability to serve as “a digital bridge.” The 24/7 monitoring provided by Didge boosts transparency, monitors safety, and expedites progress tracking. The platform is designed to be customizable and multi-functional to include capabilities such as 360° video capture and daily laser scanning. The platform also serves to monitor environmental conditions and enhance security surveillance on project sites.

Fully Autonomous Robots

The robotics platform is designed to maximize reliability and safety, while minimizing the cost of operation. Didge’s robots are fully autonomous, and the fleet does not require human supervision. Each robot can drive for up to 6 hours, and then park itself at the nearest base station for recharging. To achieve precise localization and navigation through the complex construction site environment, the robots are using an approach similar to self-driving vehicles. Nextera’s autonomous navigation software uses Deep Learning to process data from advanced depth sensors, cameras, lidars, inertial sensors integrated with the robot’s body, making real-time decisions. The Neural Networks for localization, navigation, and mapping are trained on data collected over thousands of miles driving inside active construction sites.
 width=The Didge robot fleet currently includes wheeled and tread versions, capable of scaling stairs or using human-operated hoists.  Regardless of the robot type, all are sharing the same “brains” and “eyes”: high-performance edge computers and advanced navigation software, and all are connected to a fleet management cloud.

Smart Data Enables Progress Tracking, Safety Monitoring, and More

The data collected via the robotics platform offers a multitude of uses from safety, to progress tracking, and post-completion data for building owners through 360-images delivered on a 3D map. The platform provides a visual archive giving behind the walls look into a building’s infrastructure once the project is complete.

Lana Graf, CEO and founder of Nextera Robotics, describes the potential of the deep learning powered software “to make a long-awaited shift in construction.”

Michael McKelvy, president and CEO of Gilbane Building Company framed the joint-venture partnership as “a tremendous opportunity to deliver greater value to our clients by using robotics to help us build smarter and with the highest levels of safety.”

The robots are currently on site at 225 Wyman, an innovation and life science campus project in Waltham, MA. Gilbane plans to deploy robots to sites across its divisions in the fourth quarter of 2021.

Watch our video introducing the Didge robotics platform:




About Authors
Eric Cushman leads Gilbane’s joint venture partnership with Nextera Robotics, managing the company’s efforts to develop and scale a new AI platform delivered through autonomous robots. Cushman also serves as Operations Manager for Gilbane’s Connecticut office. He is responsible for all project management, supervision, and support operations staff in the Connecticut office, as well as a principal and project executive on strategic projects. Cushman who is one of the company’s youngest operations managers, began his career at Gilbane in 2007.
Read more posts by Eric Cushman