The MARIO project is developing a team of humanoid, quadruped, and aerial robots to bring remote, real-time monitoring to complex construction sites
MARIO, or Multi-Agent Robotic system for Inspection On-site, is a multi-agent heterogenous robotic system that can enable construction inspectors to perform remote real-time construction progress monitoring on complex environments of construction job sites with multiple robotic agents capable of various locomotion.
MARIO is designed and developed at Virginia Tech’s ARCADE research lab managed by Dr Kereshmeh Afsari, associate professor of construction engineering and management in the Myers-Lawson School of Construction. MARIO is sponsored by Procon Consulting and developed in close collaboration with Stephen DeVito, director of technology at Procon Consulting.
Multi-agent robotic systems hold tremendous promise for the construction industry by enabling new levels of automation, boosting productivity and supporting large-scale, complex operations. By decomposing lengthy construction tasks into parallel, manageable sub-tasks, these systems allow multiple agents to work simultaneously, transforming traditionally sequential construction activities into efficient, high-throughput operations.
Most importantly, the demand for remote inspection and monitoring technologies in construction is steadily increasing as projects span wider geographic areas and the industry continues to face a shortage of skilled labour. The MARIO system addresses this challenge by enabling reliable remote access to construction sites through an army of integrated advanced robotic technologies providing real-time reality capture capabilities from the job site.
Innovation in construction inspection
Effective construction inspection demands access to every corner of a job site, from routine work areas to high-risk locations such as above-ceiling spaces, going up the
ladders and also confined spaces such as in crawl spaces where fall and entrapment hazards are common.
By deploying an integrated and coordinated team of robotic agents, these risks can be dramatically reduced. The humanoid robot can perform inspection tasks traditionally handled by humans with advanced vision, a quadruped robot can safely navigate confined and hard-to-reach environments with agile locomotion, and an aerial robot can conduct high-elevation inspections without exposing inspectors to fall hazards.
Together, these robotic platforms enabled with AI capabilities expand inspection reach while significantly enhancing safety and improving inspection effectiveness, enabling comprehensive monitoring that was previously difficult, dangerous or impossible.
MARIO is currently in the first phase of development. The long-term vision for MARIO is to be validated in multiple scenarios on the dynamically changing environment of the job site to allow an army of robots access construction sites safely and effectively at any point in time to provide human inspectors and project members fast real-time accurate reality capture and information about the status of construction.
Despite the adoption of advanced construction engineering and management strategies, the construction process often remains a “black box”. Its highly complex, fragmented and dynamic nature obscures clear visibility into how, when and why key decisions are made, costs escalate or safety risks emerge.
As a result, the true real-time status of a project frequently remains uncertain, leaving critical insights hidden within the very processes meant to deliver successful outcomes. Over the next five to 10 years, MARIO will evolve into an advanced scalable platform integrated that can incorporate any robotic technology on demand for real-time transparency and intelligence in construction projects. MARIO will enable demystifying the construction “black box” with real-time reality capture capability to bring transparency to previously hidden project dynamics.
Next-level AI-driven intelligence
MARIO can scale the ability of a single inspector by allowing them to supervise multiple robots across multiple sites. This allows them to review structured visual data from a single location rather than spending days traveling between locations and walking on every floor. That lets the same number of experts cover more work without lowering the quality of services.
Multi-agent systems like MARIO can also open the door to people who might not fit the traditional mould of a field inspector. If physical demands or location have been barriers, remote inspection workflows allow someone to contribute from a distance, using robots as their presence on site. That can make inspection and construction-related roles more accessible to people with disabilities, caregivers or those living far from major construction sites.
This technology can unlock the digital twins future to have a live digital representation of the physical and functional characteristics of a building that can be used to improve its lifecycle management.
In order to achieve digital twins, however, we first must create a methodology to continuously capture the reality of the ever-changing conditions of facilities and infrastructure.
So, MARIO would be always evaluating and measuring the real world through its various sensing systems. Twenty years from now, it is expected to be the norm for every building to have a real-time visual AI-powered and robot-enabled digital twin – a living image and sensor-based record that anyone responsible for or using that facility can tap into.
In that world, surprises during construction, turnover or operations should be the exception, not the rule, because the visual truth is always available and reconciled with all the other facility data we rely on.
The post MARIO: Unlocking the potential of AI and robotics in construction appeared first on Planning, Building & Construction Today.
