Graduate Student
The NIST Dimensional Metrology Group (DMG) is seeking a motivated individual with a background in
computer science, engineering or physics to join our team. The DMG has a long history in developing
test procedures to evaluate the performance of dimensional measurement systems (such as Cartesian
coordinate measuring machines, laser trackers, terrestrial laser scanners, X-ray computed tomography
systems). These test procedures are then absorbed into national and international documentary
standards providing traceability, enabling commerce, and ensuring the competitive advantage of US
industries. Developing these procedures is time-intensive because it requires identifying and testing for
error sources. As measurement instrument technologies advance, sufficiently modeling error sources
becomes increasingly intractable. For example, X-ray Computed Tomography (XCT), used in quantum
sensor development and semiconductor manufacturing, and terrestrial laser scanners (TLSs), used in
communication and robotics, are instruments whose inherent complexity makes comprehensive
modeling nearly impossible. This difficulty leads to severely protracted standards development
processes, with efforts taking years or decades to mature. We hope to radically accelerate this process
by harnessing the power of agentic-AI to identify possible error sources and testing schemes to create
the foundation for a starting document that can be further refined by a standards developing
committee. We have two objectives: 1) we will assess the feasibility of applying agentic-AI for this
purpose using an ongoing documentary standards effort for TLSs as a testbed within the NIST Forensic
Science Standards Program (through the Organization of Scientific Area Committees [OSAC] for Forensic
Science administered by NIST) and subsequently deploy the technique to XCT in an ongoing ASME/ISO
standards effort and 2) as we develop a local agentic-AI (within the RChat framework) that can run
different LLMs to find the model that produces the best results, we hope to develop protocols (such as
best practices guide or standards) for agentic-AIs tailored to supporting the development of
documentary standards. As a member of our multi-disciplinary team, you will collaborate closely with
experts in dimensional metrology, engineering, mathematics, and computer science to develop AI
models to understand error sources in measurement systems and develop draft documents that can be
considered by standard committees. You will present the work at meetings and conferences and publish
results in peer-reviewed journals.
Agentic-AI and dimensional measurement documentary standards
- A Bachelor’s degree in computer science, engineering or physics.
- Understanding error modeling and statistical analysis within a manufacturing context.
- Experience developing Agentic-AI workflows utilizing various Large Language Models (LLMs).
- Ability to design and implement AI-driven tools to automate and accelerate the creation of technical
documentary standards. - Strong oral and written communication skills.
Key responsibilities will include but are not limited to:
- Create and populate a secure database with 1) published TLS literature, 2) ASME, ISO, ASTM,
and OSAC standards, 3) available mathematical error models, 4) large quantities of 3D measurement data from TLSs along with reference data (true values), and 5) test requirements
(such as time and cost limitations). - Build and test a local agentic-AI with different LLMs to find the model that best meets our
needs. - Develop protocols that may be used in the future for how to build an agentic-AI for
documentary standards. - Present results at internal meetings, and occasional meetings with external stakeholders.
- Ensure that results, protocols, software, and documentation are archived appropriately.