Research questions at the intersection of advanced AI, robotics, evolving technologies, and quantitative evidence.
Each domain begins with a specific question, a relevant operating context, and evidence that can be meaningfully examined.
SELECT A RESEARCH AREA
Which evaluation findings still hold as AI capabilities, tools, and task environments change?
Benchmark design · Reliability claims · GeneralisationAI evaluations and consequential deployment decisions
SCOPING QUESTIONDoes the comparison address the operating conditions you intend to approve?
Examine task success, human intervention, recovery, and untested conditions before drawing conclusions about autonomy.
Autonomy claims · Human intervention · Operating envelopeHumanoid systems, embodied AI and deployment readiness
SCOPING QUESTIONWhat does the available test record establish beyond the demonstration?
Assess whether past tests remain applicable when software, sensors, physical configurations, or operating histories change.
Technology qualification · Digital twins · Evidence after modificationEngineering qualification and system-change decisions
SCOPING QUESTIONWhich part of the prior evidence remains decision-relevant after modification?
Review a material computational claim against its underlying records, alternative explanations, and limits of inference.
Claim–evidence boundary · Provenance · Comparative assessmentR&D, investment and technical diligence
SCOPING QUESTIONWould a different interpretation of this result change a consequential decision?
Cross-sector work may be considered where the question, evidence, and necessary expertise can be clearly defined.
Each proposal is assessed individually. Hardware testing, regulated evaluation, certification, and legal advice require suitably qualified specialists and separately agreed arrangements.
Share a non-confidential question, the decision it affects, and the relevant timeframe.
Discuss a technical questionWhether a team is based in China, the United States, or elsewhere, collaboration begins with a concrete technical decision and an assessable question. Data access, responsibilities, and rights to resulting work are then agreed explicitly.
Discuss a research collaboration