Standardizing Spatial Data Models and Workflows to Accelerate Innovation in XR, Robotics, and AI
ECE Assistant Professor Mallesham Dasari, in collaboration with Jacob Chakareski from New Jersey Institute of Technology, was awarded a $300K NSF grant for “POSE: Phase I: Enabling an Open-Source Ecosystem for 4D Dynamic Geometry in XR, Robotics, and Autonomous Systems.” Dynamic 3D and 4D data are becoming essential building blocks for emerging technologies such as extended reality, robotics, AI, and autonomous systems. However, the software ecosystem supporting these data remains fragmented, with incompatible tools, formats, and workflows limiting collaboration and innovation. This NSF-funded project will build on Open4D, an open-source platform developed at Northeastern University for representing, processing, and sharing dynamic 4D geometry data. The project will engage researchers, developers, and industry stakeholders to design a broader community-driven ecosystem around Open4D, including shared standards, interoperable tools, governance models, and evaluation methodologies.
Abstract Source: NSF
This Pathways to Enable Open-Source Ecosystems (POSE) project addresses the increasing demand for tools that handle fast changing 3D and 4D visual information used in robotics, extended reality, artificial intelligence, autonomous systems, and similar areas. Current resources are scattered and incompatible, slowing progress and limiting access. This project addresses these challenges by planning a shared, open ecosystem for dynamic 4D geometry data, enabling common standards, tools, and workflows that can be used across domains. The project supports broader participation in spatial computing and accelerates progress in applications such as immersive education, remote operation, and intelligent automation. This effort directly serves the national interest by strengthening the data infrastructure underlying next-generation computing systems. Advances in interoperable spatial data processing can improve the reliability and responsiveness of systems used in areas such as disaster response, critical infrastructure monitoring, and defense and surveillance operations, where timely and accurate spatial understanding is essential. At the same time, the project contributes to economic growth by enabling scalable data pipelines for autonomous systems and spatial computing platforms, supporting innovation across industry sectors.
This project scopes and designs an open-source ecosystem for 4D dynamic geometry, building on an existing research artifact as a foundation. The effort focuses on identifying shared abstractions, interoperable data representations, and system-level requirements that enable extensible development across domains such as extended reality (XR), robotics, and autonomous systems. The team will study existing 4D geometry tools, identify common needs, and outline a unified ecosystem that supports shared formats and development infrastructure. The team interviews developers and users to define requirements for interfaces, benchmarks, and governance. The project also evaluates risk, licensing needs, and technical constraints to ensure long-term sustainability. The outcome will include a blueprint for a community-driven open-source framework for dynamic geometry. In addition, the project analyzes fragmentation across current pipelines to identify opportunities to standardize data models and workflows. It examines trade-offs in performance, scalability, and cross-platform compatibility, and assess security and robustness considerations for shared infrastructure. The resulting open-source ecosystem will include a reference architecture, governance model, and evaluation methodology to guide the future development and adoption of scalable 4D geometry processing systems.