Research report · Active
DynNav: Uncertainty-Aware Risk-Sensitive Navigation
Independent robotics research · Active · 2025–Present
DynNav connects belief-aware planning, risk-sensitive route selection, returnability constraints, and safety monitoring for robots navigating previously unmapped environments.
Abstract
Develop and evaluate navigation methods that reason about map uncertainty, collision risk, returnability, and safe replanning in unknown environments.
Scientific question: How should a robot trade off path efficiency, uncertainty exposure, and returnability in unknown environments?
Contribution: A risk-aware navigation research repository connecting belief-aware planning with safety monitoring.
Marked research artifact slot
Risk-aware planner architecture
Architecture figure slot for verified diagrams of modules, data flow, and software boundaries.
Marked research artifact slot
Belief map · risk cost · safe replanning
Pipeline slot for GIFs or animations that explain estimation, planning, perception, or evaluation steps.
Marked research artifact slot
Synthetic risk-aware map
Demo slot reserved for real project videos or synthetic demos that are explicitly labeled as synthetic.
Research questions
- How can a robot plan safely when the map is incomplete or uncertain?
- Can risk-sensitive planning reduce collision or entrapment rate compared with expected-cost planning?
- How should a robot trade off path efficiency against safety and returnability?
Methods
Implementation status
Implemented
- Risk-aware navigation framing
- Planner comparison scaffold
- Safety-monitoring roadmap
Research prototype
- Uncertainty maps
- Safe-mode navigation
- Returnability constraints
Planned
- Seeded benchmark suite
- Statistical reporting
- ROS 2 simulation integration
Experiments and metrics
Experiments
- Uncertainty estimation
- Belief-space planning
- Safe-mode navigation
- Formal safety shield experiments
Metrics
- collision rate — planned
- path risk
- path length
- returnability success — planned
Quantitative benchmark tables will be added only after reproducible experiments are available.
Limitations
- Quantitative claims are intentionally withheld until reproducible benchmark runs are available.
- Visual figures and GIFs are placeholders unless a project page explicitly states that an artifact is generated from real experiments.
Reproducibility plan
Experiment workflow
Each module is prepared to document configuration files, datasets, evaluation metrics, and repeated experiment runs before quantitative claims are shown.
Repository setup
Public repositories are linked when available. Private, planned, or incomplete repositories remain clearly labelled rather than being presented as finished systems.
Literature context
- Risk-sensitive planning
- Control Barrier Functions
- Signal Temporal Logic
- Belief-space planning
Roadmap
- Report seeds and variance
- Add reproducible benchmark tables
- Expand hardware or simulation evaluation