Dissertation · 2026 · Open access
Safe and Uncertainty-Aware Decision-Making for Autonomous Vehicles
Deep Blue (University of Michigan) · Published 1 January 2026 · DOI 10.7302/dspace/30507
In brief
“Safe and Uncertainty-Aware Decision-Making for Autonomous Vehicles” is a 2026 dissertation by Xiao Li, published in Deep Blue (University of Michigan). OpenAlex counts 0 citations as of 22 November 2026. It is open access: the full text is free to read at University of Michigan Library.
- Citations
- 0
- Authors
- 1
Abstract
Autonomous vehicles have the potential to reduce traffic accidents, improve transportation efficiency, and expand mobility. However, ensuring safety in complex, real-world driving environments remains a fundamental challenge. This requires addressing problems across the full autonomy stack: making decisions and planning paths in traffic environments shared with human drivers whose intentions are not directly observable, perceiving the surrounding world through deep neural networks whose outputs may be imprecise, ensuring that low-level tracking controllers formally respect safety constraints given a planned path, and verifying and enforcing safety of the entire closed-loop system when neural network-learned components are embedded in the autonomy stack. This dissertation addresses these four interconnected challenges spanning the perception, decision-making, planning, and control modules of the autonomous driving system. First, we propose an interaction-aware decision-making framework for autonomous vehicles operating among human drivers with uncertain intentions. Drivers' interacting intentions are encoded as latent social-psychological parameters in a behavioral model grounded in Social Value Orientation and estimated online using a Bayesian filter. A Socially-Aware Neural Network based on the attention mechanism is designed to imitate the behavioral model for online deployment, and a decision tree search algorithm is proposed for efficient computation with linear scalability in the number of interacting drivers and prediction horizons. Evaluations in forced highway merging scenarios using both simulated environments and real-world traffic datasets demonstrate high prediction accuracy and successful merging across various traffic scenarios. Second, we address uncertainty quantification in deep neural network-based perception. A deep ensemble of heterogeneous neural network regressors is integrated with conformal prediction to jointly estimate vehicle states and their associated uncertainty bounds from RGB image inputs, with probabilistic guarantees under exchangeability assumptions. A Conformal Tube Model Predictive Control scheme is proposed to incorporate these uncertainty estimates and ensure safety. Evaluations in a high-fidelity simulator demonstrate the effectiveness of the proposed approach in maintaining safe following distances and accurate speed tracking, including under out-of-distribution scenarios. Third, we develop a supervisory safety filter that integrates parameterized Control Barrier Functions with a Reference Governor for safe reference tracking. The Control Barrier Function-based safety filter minimally adjusts the nominal control input to enforce safety constraints, while the Reference Governor dynamically tunes the barrier parameters online. Theoretical analysis establishes that this integration ensures persistent safety with the adjusted barrier parameters converging to the desired reference in finite time. Simulations on vehicle lateral reference tracking demonstrate reliable constraint enforcement, reduced conservatism, and minimal intervention to the nominal policy compared to benchmark safety filters. Fourth, we present algorithmic methods for synthesizing Control Invariant Sets for dynamical systems…
Cite this paper
Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.
APA 7
Li, X. (2026). Safe and Uncertainty-Aware Decision-Making for Autonomous Vehicles. Deep Blue (University of Michigan). https://doi.org/10.7302/dspace/30507
MLA 9
Li, Xiao. “Safe and Uncertainty-Aware Decision-Making for Autonomous Vehicles.” Deep Blue (University of Michigan), 2026. https://doi.org/10.7302/dspace/30507.
Chicago
Li, Xiao. 2026. “Safe and Uncertainty-Aware Decision-Making for Autonomous Vehicles.” Deep Blue (University of Michigan). https://doi.org/10.7302/dspace/30507.
Harvard
Li, X. (2026) ‘Safe and Uncertainty-Aware Decision-Making for Autonomous Vehicles’, Deep Blue (University of Michigan). doi:10.7302/dspace/30507.
IEEE
X. Li, “Safe and Uncertainty-Aware Decision-Making for Autonomous Vehicles,” Deep Blue (University of Michigan), 2026, doi: 10.7302/dspace/30507.
BibTeX
@phdthesis{li2026safe,
title = {{Safe and Uncertainty-Aware Decision-Making for Autonomous Vehicles}},
author = {Li, Xiao},
year = {2026},
doi = {10.7302/dspace/30507},
url = {https://doi.org/10.7302/dspace/30507}
}
Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)
Topic and keywords
Physical Sciences › Engineering › Automotive Engineering › Autonomous Vehicle Technology and Safety
- autonomous vehicles
- driver intention prediction
- social value orientation
- Bayesian filtering
- attention-based neural networks
- decision tree search
- highway merging
- uncertainty quantification
Work with this paper
Most-cited papers in Engineering
- A Closed‐form Equation for Predicting the Hydraulic Conductivity of Unsaturated SoilsMartinus Th. van Genuchten · Soil Science Society of America Journal · 1980 · 28,197 citations
- An improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experimentsW. C. Oliver, George M. Pharr · Journal of materials research/Pratt's guide to venture capital sources · 1992 · 27,379 citations
- Digital communicationsOwen Bishop · Routledge eBooks · 2007 · 24,890 citations
- The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysisNorden Eh Huang, Zhengwei Shen, Steven R. Long and 6 more · Proceedings of the Royal Society A Mathematical Physical and Engineering Sciences · 1998 · 24,008 citations
- Numerical Heat Transfer and Fluid FlowSuhas V. Patankar · CRC Press eBooks · 2018 · 23,491 citations
Questions about this paper
How do I cite this paper in APA style?
Li, X. (2026). Safe and Uncertainty-Aware Decision-Making for Autonomous Vehicles. Deep Blue (University of Michigan). https://doi.org/10.7302/dspace/30507
How many times has this paper been cited?
0 times, according to OpenAlex on 22 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.
Is this paper free to read?
Yes. It is open access at University of Michigan Library: sign in and open the full text from this page.
Who wrote this paper?
Xiao Li.
What is the DOI of this paper?
10.7302/dspace/30507, permanently at https://doi.org/10.7302/dspace/30507. Sign in and open it from this page.
Bibliographic data from OpenAlex (CC0), last checked 22 November 2026. ResearchGuru does not host this paper; once you are signed in, the links open it at the publisher or the repository that does.