Jerry Pratt honored with “Most Influential Paper” award

The work that put bipedal robots on their feet came from IHMC — and its key authors were honored by an international conference recently.

The groundbreaking work of IHMC’s Senior Research Scientist Dr. Jerry Pratt was honored at the 20^th International Conference for Humanoid Robots virtually in late summer of 2021.

A journal article authored by Pratt with co-authors Sergey Drakunov, Ambarish Goswami, and John Carff, an IHMC senior research associate, was named Most Influential Paper — a distinction spanning the 20-year history of the conference.

“Capture Point: A Step toward Humanoid Push Recovery” has been cited more than 1,000 times since its publication in 2007.

Today, key ideas from this project are the foundation of multiple humanoid robots’ ability to balance and recover from unexpected pushes.

The Capture Point paper has become a critical reference point for the field of bipedal robotics.

Expanding on mathematical models of balance, Pratt and his team developed formulas that allows bipedal robots to identify the area on the ground that they must step or shift to in order to avoid a fall if pushed.

The original paper was demonstrated in simulation, but soon after publication, Dr. Pratt says, IHMC researchers, collaborators, and peers in the field were able to put Capture Point ideas into practice on real robots. The IHMC team and other participants in the 2015 DARPA Robotics Challenge used Pratt’s findings to enable their robots to navigate the course.

Looking back at the significant impact this groundbreaking paper has had over the years, Dr. Pratt credits the dedication of his coworkers and colleagues and IHMC’s culture of innovation for his success and recent recognition at the 2021 Humanoids conference.

“We knew we were on to a good idea, so we did put a lot of effort into making sure it was high quality,” he says of the paper.

Pratt received the Humanoids Most Influential Paper award virtually this summer. He and colleagues continue to expand upon Capture Point concepts in the lab. He and his team are working to make IHMC’s humanoid robots lighter and more able to approximate the agility of human muscle-actuated limbs.