Led by Nikolaus Correll and his research group, CU’s first humanoid robot is being prepared to perform electric vehicle (EV) lithium-ion battery pack disassembly as part of an ARPA-E research program to create a circular supply chain for domestic EV batteries. (Photo: Caroline Harrah)

Alessandro Roncone and a PhD student are working on a collaborative robot that, in the future, will assist people across complex environments—from factories to hospitals. (Photo: College of Engineering staff)

A modified Boston Dynamics Spot robot operates in an experimental mine in Colorado in preparation for the DARPA Subterranean Challenge. (Photo: Michelle Wiese)

Artificial intelligence research advances to solve real-world challenges

CU Boulder College of Engineering and Applied Science (CEAS) researchers maneuver a robot through debris in an experimental mine tunnel in Colorado. The robot assesses the unstable ceiling before continuing toward areas too dangerous for human rescuers, searching for survivors who could be running out of air.

Other CEAS researchers are developing systems that analyze satellite imagery to identify environmental changes invisible to the human eye, while computers mine years of medical records in seconds to help doctors save lives.

These are just a few examples of how CU Boulder researchers are collaborating across disciplines to develop artificial intelligence systems that work alongside humans, addressing challenges in emergency response, space and planetary exploration, medicine, education, environmental prediction, accessibility applications and more.

When robots venture where humans cannot

Sean Humbert, professor of mechanical engineering, develops autonomous systems and bio-inspired robots for high-risk rescue operations. Testing robots in unmapped caves during the Defense Advanced Research Projects Agency (DARPA) Subterranean Challenge, Humbert collaborated with Christoffer Heckman, associate professor of computer science, and Eric Frew, professor of aerospace engineering sciences,  to uncover navigation problems impossible to predict in labs.

In space applications, where missions are costly and communication delays are unacceptable, Morteza Lahijanian, associate professor of aerospace engineering sciences, develops autonomous systems for spacecraft that must dock with satellites and space stations without human intervention. His mathematical verification methods ensure these systems can handle situations no programmer could anticipate.

Beyond recognition: AI that truly sees

Danna Gurari, assistant professor of computer science, applies neural network models to generate spatial descriptions rather than standard object classification. Developed with input from individuals who have vision impairments, her systems provide detailed image descriptions that enable AI-generated spatial navigation. People require different types of information in various contexts. Her work addresses the challenge of providing relevant details without overwhelming users with unnecessary information.

Teaching robots to work with people

Alessandro Roncone, assistant professor of computer science, designs robots that act as teammates by understanding human goals and adapting accordingly. His work addresses tricky questions: When should the robot override the human? How can it warn people about limitations without being annoying? Roncone’s research suggests that some chemists hate when robots second-guess their decisions. Others become too dependent on robotic assistance. Achieving balance requires understanding individual personalities.

Making machines understand

James H. Martin, professor of computer science and faculty fellow at the Institute of Cognitive Science, explores natural language processing systems that help artificial intelligence understand language meaning for practical applications. Despite advances with large language models, current approaches struggle with ambiguous language. Martin’s team mines electronic medical records for temporal and causal patterns to improve clinical outcomes. Additionally, Martin and his students are contributing to systems to analyze 
and improve instructional discourse in K–12 STEM classrooms.

Standardizing how we teach and visualize AI architectures

Associate Professor Tom Yeh’s AI by Hand has achieved global impact, garnering over 200,000 social media followers and demystifying complex AI concepts through hand-drawn visual explanations. Yeh, a computer scientist, invented a novel unified representation framework for neural network architectures to help learners and practitioners better understand, compare and build models and engage more meaningfully with the foundations of modern AI.

Embedding Earth’s data to improve monitoring

Esther Rolf, assistant professor of computer science, develops machine learning systems that transform satellite data into succinct, general-purpose representations of Earth. These embeddings unlock large-scale applications in environmental monitoring. Her recent research leveraged these embeddings to identify previously unmonitored mining activity across sub-Saharan Africa. A key part of her mission is making satellite imagery and machine learning accessible to all.

Looking forward

As AI systems integrate into scientific, industrial and public domains, CU Boulder is addressing both technical challenges and social dimensions, including fairness and privacy protection. 

Rather than replacing human judgment, CU Boulder researchers are designing technology that enhances human capabilities and preserves human agency.