Navigating ASD with AI

Assistant computer science professor Ngoc Nguyen is helping lead a National Science Foundation project that just received $300,000 in funding. Working with an industry partner, her team is developing tools to help people on the autism spectrum find meaningful careers.

Nguyen, in the project’s first phase, will develop an AI-driven vocational assessment tool designed to provide more accurate and efficient job connections for individuals with autism spectrum disorder (ASD).

“The overarching goal of this project is to make vocational assessments more engaging and realistic for users and learners with autism,” Nguyen said.

Nguyen’s research utilizes a combination of AI approaches and human-centered computing techniques , including reinforcement learning, cognitive modeling and immersive virtual reality, to make the job search experience more personalized and supportive for neurodiverse learners. 

Typically, these assessments rely heavily on human evaluators. The process is labor-intensive, costly and often subjective. Results fail to capture the unique strengths and interests of users with ASD, and bias or limited information about job availability can lead to inaccurate pairing. When that happens, trainers and neurodiverse users often have to start over. The process is draining, costly and time-consuming.

Nguyen is working to overcome these roadblocks. 

“[Current vocational tests] can lead to mismatched job placement. We want to build more accurate evaluations of a learner’s ability and to offer more realistic ways to explore jobs through VR scenarios,” said Nguyen.

The ultimate goal of the project is simple: save users time, money and provide more accurate direction toward meaningful careers.

One approach  she has experimented with is using virtual reality. VR offers a safe environment for users to explore different jobs and tasks. From there, AI technology can help organize data and provide real-time feedback about what tasks a person should explore next. At the end of a series of tasks, the assessment provides personalized job recommendations based on available positions and the individual's observed strengths and interests.

Nguyen was intrigued by these systems that can adapt to individual learners and the rapidly developing abilities of AI. 

“I first started with interest in technology algorithm models.” She said, “Then I started asking the question: So what — what if you have more accurate algorithms? Can that new model or that new algorithm actually help people?” 

Her research interests shifted toward human-AI interaction. This discipline of scientific inquiry considers not just the function of AI systems, but rather how humans engage with that technology and how it can support — rather than replace — human capability.

“At the end of the day, the goal of any intelligence system should be designed to work with and for humans, rather than just operate in isolation.” Nguyen said,

“This project is not about developing robots or systems per se, but we are more concerned about designing and deploying a system that can dynamically adapt and help human capabilities.”

Nguyen’s research is not only making a direct impact on the lives of individuals with ASD, but it demonstrates that new technology doesn’t have to be a source of anxiety. In the right hands — in UD’s hands — it is a tool to drive innovation, empower learners, advance communities and shape systems that don’t just work, but work for everyone.