Source author record

Lavindra de Silva

Lavindra de Silva appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

Catalog footprint

What is connected

5works
6topics
4close collaborators

Actions

Connect this record

Log in to claim

Research graph

See the researcher in context

Open full explorer

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Published work

5 published item(s)

preprint2022arXiv

HyperTensioN and Total-order Forward Decomposition optimizations

Hierarchical Task Networks (HTN) planners generate plans using a decomposition process with extra domain knowledge to guide search towards a planning task. While domain experts develop HTN descriptions, they may repeatedly describe the same preconditions, or methods that are rarely used or possible to be decomposed. By leveraging a three-stage compiler design we can easily support more language descriptions and preprocessing optimizations that when chained can greatly improve runtime efficiency in such domains. In this paper we evaluate such optimizations with the HyperTensioN HTN planner, used in the HTN IPC 2020.

preprint2014arXiv

Extending Agents by Transmitting Protocols in Open Systems

Agents in an open system communicate using interaction protocols. Suppose that we have a system of agents and that we want to add a new protocol that all (or some) agents should be able to understand. Clearly, modifying the source code for each agent implementation is not practical. A solution to this problem of upgrading an open system is to have a mechanism that allows agents to receive a description of an interaction protocol and use it. In this paper we propose a representation for protocols based on extending Petri nets. However, this is not enough: in an open system the source of a protocol may not be trusted and a protocol that is received may contain steps that are erroneous or that make confidential information public. We therefore also describe an analysis method that infers whether a protocol is safe. Finally, we give an execution model for extended Petri nets.

preprint2014arXiv

HATP: An HTN Planner for Robotics

Hierarchical Task Network (HTN) planning is a popular approach that cuts down on the classical planning search space by relying on a given hierarchical library of domain control knowledge. This provides an intuitive methodology for specifying high-level instructions on how robots and agents should perform tasks, while also giving the planner enough flexibility to choose the lower-level steps and their ordering. In this paper we present the HATP (Hierarchical Agent-based Task Planner) planning framework which extends the traditional HTN planning domain representation and semantics by making them more suitable for roboticists, and treating agents as "first class" entities in the language. The former is achieved by allowing "social rules" to be defined which specify what behaviour is acceptable/unacceptable by the agents/robots in the domain, and interleaving planning with geometric reasoning in order to validate online -with respect to a detailed geometric 3D world- the human/robot actions currently being pursued by HATP.

preprint2013arXiv

A Verifiable and Correct-by-Construction Controller for Robot Functional Levels

Autonomous robots are complex systems that require the interaction and cooperation between numerous heterogeneous software components. In recent times, robots are being increasingly used for complex and safety-critical tasks, such as exploring Mars and assisting/replacing humans. Consequently, robots are becoming critical systems that must meet safety properties, in particular, logical, temporal and real-time constraints. To this end, we present an evolution of the LAAS architecture for autonomous systems, in particular its GenoM tool. This evolution relies on the BIP component-based design framework, which has been successfully used in other domains such as embedded systems. We show how we integrate BIP into our existing methodology for developing the lowest (functional) level of robots. Particularly, we discuss the componentization of the functional level, the synthesis of an execution controller for it, and how we verify whether the resulting functional level conforms to properties such as deadlock-freedom. We also show through experimentation that the verification is feasible and usable for complex, real world robotic systems, and that the BIP-based functional levels resulting from our new methodology are, despite an overhead during execution, still practical on real world robotic platforms. Our approach has been fully implemented in the LAAS architecture, and the implementation has been used in several experiments on a real robot.

preprint2013arXiv

Towards Combining HTN Planning and Geometric Task Planning

In this paper we present an interface between a symbolic planner and a geometric task planner, which is different to a standard trajectory planner in that the former is able to perform geometric reasoning on abstract entities---tasks. We believe that this approach facilitates a more principled interface to symbolic planning, while also leaving more room for the geometric planner to make independent decisions. We show how the two planners could be interfaced, and how their planning and backtracking could be interleaved. We also provide insights for a methodology for using the combined system, and experimental results to use as a benchmark with future extensions to both the combined system, as well as to the geometric task planner.