Marcus Hoerger

About

Marcus Hoerger is a robotics researcher whose work spans autonomous navigation, decision-making under uncertainty, and motion planning for robotic systems. He is best known for his contributions to Partially Observable Markov Decision Processes (POMDPs), a mathematically principled framework that enables robots to make reliable decisions despite incomplete information about their environment. His most-cited work, "The Multilegged Autonomous eXplorer (MAX)" (2017, 23 citations), introduced an ultralight six-legged robot designed to traverse complex indoor and outdoor terrains, demonstrating his broad interest in physical robotic platforms alongside theoretical planning methods. Hoerger has made significant strides in making POMDP solvers practical for real-world deployment, developing software frameworks, online solvers for continuous observation spaces, and innovative Multilevel Monte Carlo approaches to reduce computational costs. His 2021 "POMDP-Based Candy Server" project memorably illustrated these algorithms operating over a seven-day real-world demonstration. His work on linearization and non-linearity measures further advances motion planning for systems with complex dynamics. With over 75 cumulative citations across his published research, Hoerger's contributions are shaping how autonomous robots reason and act intelligently in uncertain, real-world environments.

Research Focus

Key Achievements

Top Papers

1. 1
   The Multilegged Autonomous eXplorer (MAX)

   23 citations · 2017
2. 2
   A Software Framework for Planning Under Partial Observability

   12 citations · 2018
3. 3
   Linearization in Motion Planning under Uncertainty

   9 citations · 2020
4. 4
   POMDP-Based Candy Server:Lessons Learned from a Seven Day Demo

   9 citations · 2021
5. 5
   Multilevel Monte-Carlo for Solving POMDPs Online

   7 citations · 2022
6. 6
   Multilevel Monte Carlo for solving POMDPs on-line

   5 citations · 2022
7. 7
   A distributed, any-time robot architecture for robust manipulation

   4 citations · 2018
8. 8
   Non-linearity Measure for POMDP-based Motion Planning

   4 citations · 2024
9. 9
   An On-Line POMDP Solver for Continuous Observation Spaces

   2 citations · 2021
10. 10
    Multilevel Monte-Carlo for Solving POMDPs Online

    2 citations · 2019

Key Collaborators

Contact & Links