Dissertation · 2026 · Open access

Hybrid Architectures for Motion Planning and Control of Autonomous Systems under Uncertainty

Mohamed-Khalil Bouzidi

Published 1 January 2026 · DOI 10.17169/refubium-52504

Open full textCite this paper

In brief

“Hybrid Architectures for Motion Planning and Control of Autonomous Systems under Uncertainty” is a 2026 dissertation by Mohamed-Khalil Bouzidi. OpenAlex counts 0 citations as of 22 November 2026. It is open access: the full text is free to read at Universitätsbibliothek der FU Berlin Hochschulschriftenstelle u. Dokumentenserver.

Citations
0
Authors
1

Abstract

Autonomous vehicles must navigate in uncertain, dynamic environments where the behavior of surrounding traffic participants is inherently multi-modal and difficult to predict. Learning-based methods have shown promising performance for autonomous driving, demonstrating the ability to capture complex patterns in human behavior directly from data. They bring new capabilities in handling uncertainty through their expressiveness, learning from large-scale datasets, but simultaneously introduce new challenges such as reliability and safety guarantees. In particular, learning-based motion predictors have achieved impressive performance in capturing behavioral uncertainty by providing multi-modal predictions but their effective integration into closed-loop motion planning frameworks remains a challenge. This thesis addresses this gap by developing hybrid architectures that combine the expressive power of learning-based multi-modal predictions with the formal safety reasoning of model-based planning. First, it is demonstrated that this expressive power can be exploited to generate diverse ego trajectory proposals. These rich solution candidates are verified and refined by model predictive control, improving overall planning performance. Building on this insight, a contingency planning framework based on Branch Model Predictive Control is introduced, leveraging multi-modal predictions of the traffic agents behavior to maintain multiple planning strategies in parallel and delay commitment to one until uncertainty resolves. To maintain computational tractability and safety, scenario tree construction methods ensuring coverage of all prediction modes are proposed, alongside an adaptive decision postponing mechanism balancing information gain against conservatism. The thesis further introduces a closed-loop evaluation pipeline that integrates different motion predictors directly into the planning framework and assesses their impact on actual driving performance in realistic traffic scenarios. This pipeline reveals that open-loop motion prediction benchmark performance does not necessarily translate to better closed-loop driving performance, exposing a fundamental misalignment between current benchmarking practices and real-world autonomous driving requirements. Beyond this finding, experimental validation across diverse traffic scenarios demonstrates that the proposed hybrid architectures integrated with different state-of-the-art predictors consistently improves safety, comfort, and efficiency compared to existing planners.

Cite this paper

Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.

APA 7

Bouzidi, M.-K. (2026). Hybrid Architectures for Motion Planning and Control of Autonomous Systems under Uncertainty. https://doi.org/10.17169/refubium-52504

MLA 9

Bouzidi, Mohamed-Khalil. “Hybrid Architectures for Motion Planning and Control of Autonomous Systems under Uncertainty.” 2026. https://doi.org/10.17169/refubium-52504.

Chicago

Bouzidi, Mohamed-Khalil. 2026. “Hybrid Architectures for Motion Planning and Control of Autonomous Systems under Uncertainty.” https://doi.org/10.17169/refubium-52504.

Harvard

Bouzidi, M.-K. (2026) ‘Hybrid Architectures for Motion Planning and Control of Autonomous Systems under Uncertainty’. doi:10.17169/refubium-52504.

IEEE

M.-K. Bouzidi, “Hybrid Architectures for Motion Planning and Control of Autonomous Systems under Uncertainty,” 2026, doi: 10.17169/refubium-52504.

BibTeX

@phdthesis{bouzidi2026hybrid,
  title     = {{Hybrid Architectures for Motion Planning and Control of Autonomous Systems under Uncertainty}},
  author    = {Bouzidi, Mohamed-Khalil},
  year      = {2026},
  doi       = {10.17169/refubium-52504},
  url       = {https://doi.org/10.17169/refubium-52504}
}

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 driving
  • motion planning
  • motion prediction
  • hybrid architecture
  • model predictive control
  • contingency planning
  • behavioral uncertainty
  • closed-loop evaluation

Work with this paper

Most-cited papers in Engineering

All highly cited Engineering papers

Questions about this paper

How do I cite this paper in APA style?

Bouzidi, M.-K. (2026). Hybrid Architectures for Motion Planning and Control of Autonomous Systems under Uncertainty. https://doi.org/10.17169/refubium-52504

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 Universitätsbibliothek der FU Berlin Hochschulschriftenstelle u. Dokumentenserver: sign in and open the full text from this page.

Who wrote this paper?

Mohamed-Khalil Bouzidi.

What is the DOI of this paper?

10.17169/refubium-52504, permanently at https://doi.org/10.17169/refubium-52504. 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.