Vertically Integrated Projects (VIP) provides UMD students of all backgrounds, disciplines, and levels of experience with the opportunity to engage in scaffolded, multidisciplinary team-based projects.
VIP Team Information
VIP Team
Low-Cost Robot Scientists
VIP Faculty Mentors:
Dr. Gilad Kusne
Department of Electrical and Computer Engineering
Meeting Location, Day & Time:
Time: Tuesdays 2-3:30pm
Location: Chemistry Building, Room 1147. This is the HERBIE lab, with benchtop access; weekly meetings and hands-on work both happen there.
Description:
What if every classroom in the world could have its own robot scientist? Self-driving labs, where robots run experiments and AI picks what to try next, are changing how new materials, medicines, and electronics are discovered. But commercial systems cost tens to hundreds of thousands of dollars, which shuts out most schools and communities. HERBIE is our answer: a robot that costs about $500, built from 3D-printed parts, a Raspberry Pi, small motors, and sensors. It can mix liquids, take measurements, and use AI to decide which experiment to run next, all on its own.
HERBIE and its predecessor have been used in UMD courses since 2020, demonstrated to policymakers at a U.S. Senate Robotics Day event and a White House OSTP event on AI, and adopted by teams at other universities, national labs, and industry. Tested versions of the design are released openly, so schools and labs, especially in under-resourced and developing communities, can build HERBIE and use it to do real research.
You'll join one of four subteams, working hands-on in our lab:
Electronics: Test the first version of HERBIE's custom motor-driver circuit board, find what needs fixing, and design the next version.
Mechanical: Improve HERBIE's 3D-printed frame, liquid handling, and assembly so anyone can build one.
Software & AI: Write the code that runs the robot and the machine learning that decides which experiment to run next.
Science & Curriculum: Design experiments HERBIE can run on its own, and turn them into lessons for classrooms around the world.
Methods:
CAD and 3D printing; mechanical design and assembly
Circuit board testing, debugging, and PCB design; sensors and stepper motor control
Python programming on Raspberry Pi; version control and collaborative engineering
Machine learning for science: Gaussian processes, active learning, Bayesian experiment design
Designing and running automated chemistry and materials experiments
Technical documentation, assembly guides, and teaching materials
Majors:
All majors welcome. Especially relevant: Electrical, Computer, Mechanical, Materials Science, Chemical & Biomolecular, Bioengineering, and Aerospace Engineering; Computer Science; Mathematics; Physics; Chemistry & Biochemistry; Information Science; Immersive Media Design; Education.
Preferred Interests:
Curiosity about robots, AI, or how science gets done. No prior experience is required: you'll learn by building. Experience with CAD, 3D printing, soldering, Arduino or Raspberry Pi, Python, or circuit testing helps but is not required. We value teammates who document their work clearly and share with others. The team works in a chemistry lab, so all members complete UMD lab safety training before starting bench work. A commitment of three or more semesters is encouraged, so you can grow from new member to subteam lead.