Jiaying [Vina] wei
Research Project Threads
A closer look at the key projects that define my academic journey, bridging the gap between material physicality and computation.
Orchestrating Heterogeneous Multi-Robot Construction A Human Factored Collaboration Framework
Status:
In Development
Investigation Area:
Multi-Robot System; Centralized Planning; Human-Engaged Robotic Construction; UGV; Mobile Robots; Assembly
Stage:
PhD Proposal
The future of construction robotics will be shaped by heterogeneous robots that share dense, evolving construction sites while collaborating with human workers. Yet the research domains that ground this cross-disciplinary study remain loosely connected. Current construction robotics deployments still focus mostly on isolated work cells, relying heavily on arbitrary protection separation distances (PSD) that discourage close-proximity collaboration, while multi-robot planners consider less human intrusion, treating humans only as moving obstacles rather than decentralized agents with explicit rights and intentions. This thesis proposal states: The integration of heterogeneous robot fleets in future architectural construction requires systems to treat on-site humans as first-class participants in the workflow, by scrutinizing their communication and interventions as explicit spatio-temporal constraints. So that task feasibility and production-level efficiency can be achieved simultaneously while complying with human safety. It proposes delivering an MRS TAMP Orchestration framework that adheres to the Right-of-Way Policy (RoW), which defines yielding principles for active construction sites. It echoes Asimov’s” laws of robotics” as a literary archetype of moral agency in human priority. In this framework, the planning stack and tasks are configured differently for heavy industrial robots and mobile assistive manipulators, each tailored to its respective advantages.
The framework will be investigated through two parallel tracks: 1. Physical prototyping (WG-A): a full-scale frame structure prototype with secondary finishes to understand empirical parameters for the framework. A large industrial arm assembles primary structural elements, and a mobile robot performs secondary or surface finishing, while human workers assist with dexterous tasks and introduce natural perturbations that ask robots to yield and replan. 2. Scalable simulation (WG-B) track that explores a broader range of layouts, task graphs, and disturbance patterns. In evaluation, safety is assessed through separation statistics and RoWC compliance. Productivity and flow are measured quantitatively by prototype execution, conflict rates, and throughputs. The thesis aims to make three long-term contributions. First, it will provide a physical implementation of cross-scale studies of multi-agent collaboration in frame assembly and finishing. Second, it delivers a transferable Right-of-Way Orchestration for multiple robot types with different hierarchies and tolerance to human perturbation. Third, it provides a human-aware autonomy that is both ethically sound and practical. In doing so, the thesis sketches a general template for how robots should move through human environments in other domains as well: coordinated with one another, guided by structured task graphs, and always comply to human Right-of-Way.
see full proposal at https://www.architecture.cmu.edu/events/phd-cd-doctoral-proposal-presentation-vina-wei
Responsive Robotic Assembly with Heterogeneous Materials
Status:
Finished
Investigation Area:
Robotic Vision; Robotics Assembly; Human Robot Collaboration
Category:
Master thesis
advisor:
Prof. Joshua Bard (CMU)
Prof. Daniel Cardoso Llach
A Inquiry into Robot visions...TBA
Rethinking Automation in Construction
Status:
Finished
Investigation Area:
Human assistive Robot; UGV; On-site construction; reinforcement learning; human tracking
Funding:
Open Access funding provided by Carnegie Mellon University. This work received support from the Commonwealth of Pennsylvania through its PA Department of Community and Economic Development, and the Manufacturing Futures Institute at Carnegie Mellon University. An A+MI Award from Google and a gift from Autodesk to the corresponding author provided additional support.
PI:
Daniel Cardoso Llach (SOA)
Jean Oh (RI)
Combining ethnographic and technical design research methods, this research contributes a path towards the human-centered development of construction robotics technologies. Since the 1980s, researchers have envisioned robots performing skilled construction tasks such as bricklaying, painting, spray-coating, or site monitoring. In this article, we envision instead robots performing simpler, ancillary activities supporting workers while they collaboratively carry out building tasks. We draw from an extended ethnographic engagement with construction workers to inform the design of a prototype rover able to accompany and deliver tools to carpentry workers installing formwork panels. Following a review of the state of the art in construction robotics and ethnography in technology design, we show how insights drawn from our ethnographic study informed the robot’s design as well as its innovative deep reinforcement learning (DRL) architecture for social navigation. Evaluating the robot in simulations, lab settings, and on a construction floor we document its benefits, including apt social navigation and user comfort in construction floors, and reflect on its limitations. Proposing “robot in the loop” as a design pattern combining ethnographic and technical design research, the article shows how the world of construction might be brought closer to the world of technology design, centering workers’ contexts and experiences in the design of new technologies aimed at supporting them.