InteLiPlan: An Interactive Lightweight LLM-Based Planner for Domestic Robot Autonomy


Kim Tien Ly1, Kai Lu2, Ioannis Havoutis1
1 K. T. Ly and I. Havoutis are with the Oxford Robotics Institute, University of Oxford, Oxford, UK.
2 K. Lu is with the Department of Computer Science, University of Oxford, Oxford, UK.

            Paper Link               Code


We introduce an interactive LLM-based framework designed to enhance the autonomy and robustness of domestic robots, targeting embodied intelligence. Our approach reduces reliance on large-scale data and incorporates a robot-agnostic pipeline that embodies an LLM. Our framework, InteLiPlan, ensures that the LLM's decision-making capabilities are effectively aligned with robotic functions, enhancing operational robustness and adaptability, while our human-in-the-loop mechanism allows for real-time human intervention when user instruction is required. We evaluate our method in both simulation and on the real Toyota Human Support Robot (HSR). Our method achieves a 93% success rate in the `fetch me' task completion with failure recovery, highlighting its capability in both failure reasoning and task planning. InteLiPlan achieves comparable performance to state-of-the-art large-scale LLM-based robotics planners, while using only real-time onboard computing.


System overview


Snapshots of InteLiPlan results


Multimodal data structure

The fine-tuning data is structured as text-only data, allowing deployment on different robotic platforms without adjustments. Below is a demonstration of the vision and feasibility module as chatbots.


Simulation Experiments

We initially tested the 'fetch me' task in simulation with a Toyota Human Support robot. The results shows that InterPlan can plan the 'fetch me' action sequence, and replan upon failures with human intervention. The vision and feasibility modules in the form of multimodal perception allow LLM-based generated plans to be executable on the robot.

Planning without failures

Replanning with failures

In domestic environment

Sequential failure recovery


Real-world Experiments

In the real robot experiments, we demonstrate the InteLiPlan capabilities of providing interactive on-board planner with failure reasoning and failure recovery.

Planning without failures

This set of experiments showcases the ability of the fine-tuned model to plan the correct sequence of actions for different domestic tasks with no-failure cases.

Understanding natural language

Planning mobile manipulation tasks

Planning 1-step action sequence

Planning multi-step action sequence


Replanning with failures

The experiments present that the planner can reason around failure and inform the user to ask for guidance. Our interactive approach allows the user to provide the robot with guidance for failure recovery.

Resolving task clarification

Replanning with human guidance

Towards interactive robotic autonomy with multi-step recovery


BibTex

@misc{inteliplan,
    title = {InteLiPlan: An Interactive Lightweight LLM-Based Planner for Domestic Robot Autonomy},
    author = {Kim Tien Ly, Kai Lu, Ioannis Havoutis},
    year = {2025},
    eprint = {2409.14506},
    archivePrefix = {arXiv},
    primaryClass = {cs.RO},
    url = {https://arxiv.org/abs/2409.14506}, 
}

Contact

Have any questions, please feel free to contact Kim Tien Ly


January, 2025
Copyright © Kim Tien Ly