research news

Moog gift supports research into safer autonomous vehicles

Two excavators carry a log towards a team of researchers.

The Moog gift to UB includes the use of two mini-excavators that researchers at UB's Center for Embodied Autonomy and Robotics (CEAR) can use to demonstrate some of their ongoing work. The yellow figure at right is UB's robot dog.

By PETER MURPHY

Published July 18, 2024

Print
“The impact of this work goes beyond technological development, creating opportunities for students and faculty in our school and experts at Moog to share resources and knowledge with each other. ”
Kemper Lewis, dean
School of Engineering and Applied Sciences

Moog Inc., a global design, manufacturer and integrator of precision control components and systems, and a corporate partner of the School of Engineering and Applied Sciences, will provide $750,000 to support research in the UB Center for Embodied Autonomy and Robotics (CEAR).

The gift will fund work led by CEAR co-director Karthik Dantu, associate professor in computer science and engineering. Dantu and a team of UB faculty, students and Moog engineers will work with autonomous excavators. In addition to the gift, Moog has made two mini-excavators available for experimentation.

“This is completely a research endeavor,” Dantu says. “We will use the excavator platform as a testbed to demonstrate some of the ongoing research. Separately, Moog uses excavators for all sorts of applications.”

One of Dantu’s research areas is perception — how robots see the world. Working with the excavators, he will develop a method to give them 360-degree visibility and the ability to recognize certain objects of a certain size. These capabilities give the excavators safe perception and the ability to locate a traversable path, recognize obstacles that might be in the way and, eventually, the ability to work with other excavators and autonomous robots.

Perception also plays a significant role when the excavators dig trenches, clear debris and transport materials. The excavators must identify their target and perform these tasks with precision, carry heavy objects to an endpoint or complete other tasks typical of some of the industries that generally employ the heavy-duty machines.

Dantu says that developing guaranteed recognition, correct follow-through on certain tasks and safety for autonomous excavators could transform entire industries like construction and mining.

“In Canada, there’s a lot of mining and that domain is interesting because they’re in harsh conditions,” Dantu says. “You don’t want people to be working in minus 20 or minus 40-degree temperatures. If something is autonomous, or even if you can do remote operation, that goes a long way.”

Collaborating with Moog

This new gift is the latest step in a history of collaboration between the School of Engineering and Applied Sciences and Moog, and just one of the many ways that Moog has supported UB through the Boldly Buffalo campaign. The company established and continues to support the Moog Professor of Innovation professorship, currently held by John Crassidis, SUNY Distinguished Professor of mechanical and aerospace engineering. 

The research collaborations between the organizations thus far have produced groundbreaking developments in additive manufacturing, artificial intelligence and agricultural autonomy.

“The focus of the gift will be on technological development: to investigate and learn more about how to leverage advances in computer science relating to safe, collaborative robotic control and autonomy; to discover and develop critical pathways leading to new capabilities and uses for land, air and space vehicles, including autonomous vehicles and collaborative robotics,” says George Small, chief technology officer at Moog. “In addition to vehicles themselves, inquires of interest may extend to the systems operating in and controlling and manipulating their operating environments.”

Small says the learning associated with this work will foster collaborations between UB students and faculty, and Moog personnel. Moog experts will bring their skills and experience to the UB campus as well. They will also collaborate around architectures and solutions associated with the work, and Moog will perform real-world testing of the technology in representative environments.

Kemper Lewis, dean of the School of Engineering and Applied Sciences, will work with Dantu to support mutual engagement of faculty members and students with employees and leadership in Moog.

“This collaboration between Moog and UB has the potential to advance automation and robotics in many critical areas,” says Lewis. “The impact of this work goes beyond technological development, creating opportunities for students and faculty in our school and experts at Moog to share resources and knowledge with each other.”

Faculty members with backgrounds spanning across disciplines and students will contribute to the project funded by this gift.

Dantu is focused on perception and coordination — how robots work together — but other aspects of the project could require additional areas of expertise from faculty members in departments across the school, including:

Department of Mechanical and Aerospace Engineering

  • Souma Chowdhury, co-director of CEAR and associate professor, is working on machine learning and how robots learn different behaviors and group behaviors in swarm systems, swarm-bots and autonomous systems.
  • Ehsan Esfahani, associate professor, focuses on human-machine interactions, bipedal locomotion for both humans and humanoids, and biorobotics and mechatronics.
  • Chaozhe He, assistant professor, works with connected autonomous vehicles and is solving various problems in logistics, connectivity and control.
  • Tarunraj Singh, professor, works in optimal control and can designate specific, precise movements for robots.
  • Ryan St. Pierre, assistant professor, designs mechanisms inspired by biology.

Department of Computer Science and Engineering

  • Roshan Ayyalasomayajula, assistant professor, works with sensors and can enable robots and autonomous vehicles with radio frequency sensors to determine how dense soil is or what is on the other side of a wall.
  • Chen Wang, assistant professor, focuses on geometric learning, which takes perception a step further and allows robots to see the world the way humans do.