Seiji Tokura

Dynagent: an incremental forward-chaining HTN planning agent in dynamic domains

HTN planning, especially forward-chaining HTN planning, is becoming important in the areas of agents and robotics, which have to deal with the dynamically changing world. Therefore, replanning in “forward-chaining” HTN planning has become an important subject for future study. This paper presents the new agent algorithm that integrates forward-chaining HTN planning, execution, belief updates, and plan modifications. Also, through combination with an A*-like heuristic search strategy, we show that our agent algorithm is effective for the replanning problem of museum tour guide robots, which is similar to the replanning problem of a traveling salesman.

Background sensing control for planning agents working in the real world

Online planning agents can adapt to the dynamic environment by continuously modifying plans during the plan execution. This adaptability of the online planning agents is based on the assumption that they can recognise the events happening in the environment. However, robotic agents working in the real environment cannot obtain necessary information from the outer world by simply turning all sensors on while executing a plan. They have to actively sense the world by changing the direction of the sensor to the target, analysing the raw sensor data, extracting important information and recognising the situation. Online planning is useless if they cannot obtain necessary information from the world. Therefore, it is crucial to decide what events the agent needs to recognise in the background while executing a plan. To solve this problem, this paper introduces a new background sensing control technique by which planning agents can effectively observe the real environment and obtain important information when necessary during the plan execution.

Mobile robot self-localization based on tracked scale and rotation invariant feature points by using an omnidirectional camera

Self-localization is important for mobile robots in order to move accurately, and many works use an omnidirectional camera for self-localization. However, it is difficult to realize fast and accurate self-localization by using only one omnidirectional camera without any calibration. For its realization, we use “tracked scale and rotation invariant feature points” that are regarded as landmarks. These landmarks can be tracked and do not change for a “long” time. In a landmark selection phase, robots detect the feature points by using both a fast tracking method and a slow “Speed Up Robust Features (SURF)” method. After detection, robots select landmarks from among detected feature points by using Support Vector Machine (SVM) trained by feature vectors based on observation positions. In a self-localization phase, robots detect landmarks while switching detection methods dynamically based on a tracking error criterion that is calculated easily even in the uncalibrated omnidirectional image. We performed experiments in an approximately 10 [m] × 10 [m] mock supermarket by using a navigation robot ApriTau™ that had an omnidirectional camera on its top. The results showed that ApriTau™ could localize 2.9 times faster and 4.2 times more accurately by using the developed method than by using only the SURF method. The results also showed that ApriTau™ could arrive at a goal within a 3 [cm] error from various initial positions at the mock supermarket.

Robotic transportation system for shopping support services

A robotic transportation system for shopping support services has been developed. The system is required to act according to the customers request and position to provide services. In collaboration with mobile robots, the system provides shopping support services.

Towards Real-World HTN Planning Agents

HTN planning, especially forward-chaining HTN planning, is becoming important in the areas of agents and robotics, which have to deal with the dynamically changing world. Therefore, replanning in “forward-chaining” HTN planning has become an important subject. This chapter first presents an online planning agent algorithm that integrates forward-chaining HTN planning, execution, belief updates, and plan modifications. In stratified multi-agent planning, the parent planning agent and its child planning agents work together to achieve a goal. The parent planning agent executes a rough plan for a goal, and the child planning agents execute detailed plans for subgoals. Although this stratified multi-agent planning is efficient, it is difficult for the parent planning agent to change the plan while a child planning agent is working. This chapter also shows how to realize online planning in stratified multi-agent planning.

Emergency HTN Planning

Integration of deliberation and reaction has been an important research topic concerning agents in view of the need for an agent to react tentatively and immediately to the changing world when unexpected events occur while executing a plan. An agent is not supposed to think for a long time before reacting. Also, its reaction is not supposed to change the world greatly. However, there are some cases where deliberation is necessary for achieving an emergency goal or where the emergency plan execution prevents the resumption of the suspended plan execution. This chapter presents a new concept of on-line interruption planning that integrates deliberation and emergency deliberation. When an emergency goal is given while executing a plan, our agents suspend the current plan execution, make and execute an emergency plan, and resume the suspended plan execution. Because our agents continuously modify the suspended plans while executing an emergency plan, they can resume the suspended plans correctly and efficiently even if the world has changed greatly due to the emergency plan execution.