RoboBoom
IEEE Robotics and Automation Society· February 10, 2026· 15:58

IROS 2025 Keynotes - Human Robot Interaction session: Jing Xiao

Jing Xiao, a robotics researcher with the IEEE Robotics and Automation Society, discusses assembly collaboration, robotic nursing assistance, and social interaction for dementia support at the IROS 2025 Human Robot Interaction session.

In this talk

  1. 0:50

    Guiding robots through assembly

    Xiao describes a system where a worker uses a tracked Vive device to guide a robot with broad gestures. The robot maps those movements to precise actions, then handles tasks such as picking up parts and completing tight-clearance assembly. A single round of guidance can set up later autonomous batch work with parts in different locations and orientations.

  2. 5:39

    Teleoperated nursing assistance

    Work from Jane Lee’s lab explores a mobile nursing-assistant platform with two seven-degree-of-freedom Kinova robots, remotely operated to help patients with tasks such as retrieving a pill bottle. The team developed gaze-based and head-controlled camera views, including an eye-in-hand camera for close-up detail. Shared autonomy and automated camera control can reduce the operator’s workload, while augmented-reality cues can highlight obstacles and suggest paths.

  3. 11:01

    Measuring workload and providing feedback

    The nursing-assistance work estimates physical and cognitive workload, as well as operator intention, using wearable sensors, gaze tracking, and vision. Xiao emphasizes the role of user studies in judging whether robotic assistance works for people. A soft robotic glove developed by a colleague provides tactile and force feedback during manipulation.

  4. 12:25

    Supporting people with dementia

    Pay Fune Fiona Yuan’s lab studies robot assistance for daily activities, adjusting the amount of help to a person’s dementia stage. For milder needs, a robot may use conversation and gaze tracking to direct attention; for more severe needs, it can add gestures, object detection, and navigation. In an example, the robot either verbally guides someone to a pill bottle or leads them toward it while offering reminders.

  5. 14:50

    Choosing how much autonomy to use

    Xiao summarizes three patterns: a person guides a robot that performs manipulation, a robot supports a person doing the manipulation, and a robot guides a person through a task. She says effective interaction depends on making autonomy available at the level each task and user needs. User studies and robot learning can help systems understand intentions and choose that level of support.

Briefs summarise talks and interviews; timestamps open the video at that moment. Claims in a talk are the speaker's own.