Nearly two decades after its publication, a research paper on how moving systems avoid collisions continues to influence fields ranging from robotics to computer-aided design to digital gaming. That work by Ming Lin, Distinguished University Professor of computer science, and Dinesh Manocha, Distinguished University Professor of computer science with an appointment in electrical and computer engineering, along with their jointly advised postdoctoral researcher Jur van den Berg, has received a major award in the robotics community.

Their paper, “Reciprocal Velocity Obstacles for Real-Time Multi-Agent Navigation,” has been selected by the IEEE Robotics and Automation Society Awards Committee as the Most Influential Paper from the IEEE International Conference on Robotics and Automation (ICRA) for the 2004–2008 period. The award will be presented at the conference’s award ceremony in Vienna, Austria, in June.

The selection is based on the paper’s long-term impact among more than 3,800 papers published during that period at ICRA, a leading conference in robotics and automation. The award recognizes research that has demonstrated sustained influence and pivotal significance within and outside of the field over nearly two decades. 

The award-winning paper introduced a collision-avoidance algorithm known as “Reciprocal Velocity Obstacles”, or RVO. The method models how individual agents, such as robots or people, adjust their movements by anticipating that others will also act to avoid collisions. By incorporating this assumption, the algorithm leverages geometric reasoning and optimization-based velocity selection to enable agents to navigate shared spaces efficiently without direct communication. Many of Lin and Manocha’s graduate students extended the original RVO algorithm and developed widely used variants, including AVO, BRVO, GVO, HRVO and ORCA. 

The researchers demonstrated that the approach produces stable, smooth motion by reducing the need for repeated adjustments that can disrupt movement. It can also scale to environments with thousands of moving agents, including both static and dynamic obstacles.

They said the algorithm’s ability to account for surrounding movement is central to its practical use.

“RVO ensures that a moving robot or agent can avoid collisions with its environment and other agents, including people, animals and other robots, helping improve safety for both the system and those nearby,” they stated.

Since its introduction at ICRA in 2008, the RVO framework has remained in use across academia and industry worldwide, supporting real-time navigation in complex environments with multiple moving agents. The resulting open-source software has been downloaded by tens of thousands of researchers and developers and integrated into robotics software libraries, including ROS, as well as game engines, crowd simulation systems and animation libraries.

Lin and Manocha said the award reflects the continued relevance of their research.

“We feel a profound sense of validation that our work influenced the research landscape for many years after its publication,” they said. “We’re honored by the recognition and motivated to continue pursuing research with broader impact.” 

—Story by Samuel Malede Zewdu, CS Communications