Exploring the Tolman-Eichenbaum
Machine for Modeling Spatial Understanding in Robots

Abstract

Spatial understanding is a crucial cognitive component in all agents acting in 3D physical space, including people and robots.

We investigate the Tolman-Eichenbaum Machine (a unifying mechanistic framework for the hippocampal role in spatial and non-spatial tasks) [1] for use in biologically inspired Simultaneous Localization and Mapping (SLAM) in robots.

Establishing the preliminary context of neuroscience, we provide an approachable overview to the model and its capabilities.

We compare it to other approaches used in previous bio-inspired SLAM systems and address their limitations.

We summarize advantages and limitations of TEM and briefly outline our future work regarding its use in cognitive robotics.

References

• [1] Whittington, J. C., Muller, T. H., Mark, S., Chen, G., Barry, C., Burgess, N. and Behrens, T. E. (2020). The Tolman-Eichenbaum machine: unifying space and relational memory through generalization in the hippocampal formation. Cell, 183(5):1249–1263.
• [2] Oriolo, G., Ulivi, G. and Vendittelli, M. (1998). Real-time map building and navigation for autonomous robots in unknown environments. IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernet-ics), 28(3):316–333.
• [3] Sun, Y., Liu, M. and Meng, M. Q.-H. (2017). Improving RGB-D SLAM in dynamic environments: A motion removal approach. Robotics and Autonomous Systems, 89:110–122.
• [4] Nahavandi, S., Alizadehsani, R., Nahavandi, D., Mohamed, S., Mohajer, N., Rokonuzzaman, M. and Hossain, I. (2025). A comprehensive review on autonomous navigation. ACM Computing Surveys, 57(9):1–67.
• [5] Eichenbaum, H. and Cohen, N. (2014). Can we reconcile the declarative memory and spatial navigation views on hippocampal function? Neuron, 83(4):764– 770.

Work on the project

Introduction

    Navigation of robotic agents in real-world environments is a complex task that has been actively researched for decades. In particular, we focus on simultaneous localization and mapping (SLAM). The main challenge with SLAM has always been the spatio-temporal complexity of the environment. Navigation in static environments has been computationally solved in the 1990's [2], however as the environments get more dynamic and complex, traditional methods fall short of achieving efficient solutions. Current visual SLAM methods fail to achieve real-time performance [3] and leveraging semantic information to develop semantic SLAM remains an active area of research [4].

    In our work, we take a different approach. It is known that hippocampus in mammal brains is essential for survey navigation and evidence points towards its involvement in other types of navigation where location memory is essential [5]. The Tolman-Eichenbaum Machine (TEM) is a machine learning-based computational model of hippocampus and entorhinal cortex, formally linking them together [1]. The authors demonstrated that TEM cells resemble different cell types also found in organic hippocampus (grid cells, place cells) and learns to path integrate in unknown environments.

Timeline

• 20. 10. 2025 - START OF THE PROJECT.

  First meeting with guiding teacher, expressing interest in CogSci, picking the right topic (not too advanced, but integrated enough to classify as year project)

• 24. 10. 2025 - ESTABLISHED TOPIC, START OF RESEARCH.

  First readings of the paper. Filling out gaps in knowledge with the help of guiding teacher. Gathering resources and studying.

• 10. 11. 2025 - ESTABLISHED PROJECT WEBSITE.

  Creating a HTML website supposed to provide insight into the project's status, direction and work.

• 17. 11. 2025 - CONTINUED RESEARCH, MORE MEETINGS.

  Gaining confidence in the subject, frequent meetings with the guiding teacher in order to stay on the right track. Looking into more technical aspect of the project, planning development phase.

• 23. 1. 2026 - PROJECT HALF-YEAR REPORT.

  [ PROJECT REPORT ]

• 5. 3. 2026 - GOAL SET: PUBLISHING AT A CONFERENCE.

  As suggested by the guiding teacher, we've decided to pursue the publishing and presentation of our project at a CogSci conference in Czechia.

• 6. 3. 2026 - START OF WRITING RESEARCH PAPER, CONTINUED RESEARCH.

  Abstract, context and motivation, introduction of our work, it's relevance, importance and outlining future work. Continued research and aggregation of information and references.

• 1. 5. 2026 - FINAL DRAFT OF THE ARTICLE.

  Final touches, meetings with guiding teacher, internal reviewing and polishing. Submitting the final version to the conference organizers.
  [ ARTICLE PDF ]

• 22. 5. 2026 - FEEDBACK FROM THE ORGANIZERS.

  Our article has been accepted and reviewed by the conference reviewers. Subsequently, we received a document containing written, constructive feedback of our work and the article itself.

• 23. 5. 2026 - FINISHED PREPARATIONS FOR PRESENTATION.

  A concise, readable summary of our article in form of an A1 poster was created and printed for presentation purposes, styled in the representative look of our university.
  [ POSTER PDF ]

• 26. 5. 2026 - CONFERENCE PRESENTATION.

  Using the poster, we presented the project at a Czechoslovak conference in Czechia and let conference attendees engage with and provide feedback on our work. The presentation was a success from both our view and the view of our guiding teacher.

• 18. 6. 2026 - PROJECT HALF-YEAR REPORT.

  [ PROJECT REPORT ]