Engineers get $650K NSF grant to improve safety on construction sites

UNIVERSITY PARK, Pa. — A system for improving the safety and efficiency of construction equipment is the aim of a three-year, $650,000 grant from the National Science Foundation to a team of researchers from Penn State and the University of Illinois at Urbana-Champaign. The project will initially focus on mobile cranes, which have recently been involved in numerous construction site accidents.

Work on the project, "Safe and Efficient Cyber-Physical Operation System for Construction Equipment," will begin Jan. 1, 2016.

"Construction is a $950 billion industry in the United States and operating equipment represents one of the most dangerous tasks on a construction site. In 2013 alone, 4,045 people died as a result of construction site accidents," said Chimay Anumba, head and professor of architectural engineering at Penn State.

He and his colleagues plan to develop a system for construction equipment that combines cyber and physical components that will interact for maximum effectiveness. The system's physical components will include an omnidirectional camera to provide equipment operators with a better perspective on their work environment by detecting workers, tools and other objects on site, and sensors to collect data that will enable monitoring of conditions such as equipment-soil interaction, equipment stability, and safety of material pickup.

The researchers will develop a virtual model of the construction site augmented with live video feeds that track changes to the construction environment in real time.

"One thing that makes a site unsafe is that constantly changing and dynamic environments can create problems. For instance, traffic flows change as the work goes on. Providing equipment operators with updates in real time will have huge safety implications," Anumba said.

One scenario in which the physical and virtual components of the system will work together is when a crane operator is doing a "blind lift."

"This is when an operator can't see what he or she is lifting, so workers on the ground verbally direct them to the position of the load. By integrating cameras with the virtual model of the environment, the operator can have a much better understanding of what they are lifting in relation to other physical components on site such as workers, other equipment, materials, and power lines," said Anumba.

An intelligent reasoning mechanism will provide bi-directional feedback between the cyber environment and the physical construction site, which will then be passed along to the equipment and the operator.

Control feedback to the equipment will include restricting applied forces and torques to computed bounds — similar to anti-lock brakes — and in dire situations overriding the equipment's operation, causing the machine to stop. Feedback to the operator will consist of information about any adjustments he or she might need to make to minimize safety hazards and avoid a dangerous situation.

Part of the team's research will involve determining how to integrate all these components — a camera, sensors, virtual models of the construction site and the intelligent reasoning mechanism — into an effective system.

The team opted to use mobile cranes as the initial focus of their research.

"Among commonly used equipment, cranes account for a larger share of accidents. In crane-related accidents, the most frequent cause of worker injury or death is crane collapses at 39 percent, followed by contact with overhead power lines at 14 percent, impacts by crane loads at 14 percent, and impacts by other crane parts at 14 percent," Anumba said.

The team is collaborating with representatives from Manitowoc Cranes and HIGH Steel Structures Inc., who will provide access to cranes and to project sites for the research.

Anumba said once the team has tested and developed their system, the next step will be determining the best way for it to deliver the information to the operator.

"One idea we have proposed is creating a 'transparent cockpit' that provides control feedback to the operator in the form of both haptic cues, such as vibrations, and visual cues," he said. "Construction accidents not only result in injuries and death, but they also cause considerable delays and disruptions, negatively impacting the efficiency of operations. Our system will give operators a better understanding of their work environment, potentially resulting in safer and more efficient operations."

The team is also collaborating with two additional organizations — the National Institute of Standards and Technology construction metrology group, focusing on reliable monitoring and data collection for construction applications, and the Association of Equipment Manufacturers, an independent trade association affiliated with the material handling industry.

In addition to Anumba, the research team includes John Messner, Charles and Elinor Matts Professor in the Department of Architectural Engineering, and Penn State and University of Illinois at Urbana-Champaign researchers Mani Golparvar-Fard, assistant professor of civil and environmental engineering and computer science, and Timothy Bretl, associate professor of aerospace engineering and a research associate professor in the Coordinated Science Laboratory.

Last Updated October 14, 2015