分野別セミナー

Industrial robots have long contributed to mass production in factories, such as for automobile manufacturing. Thus, the planning and scheduling for their manipulations was given in a fixed sequential manner depending on each
product and normally designed by a specific human programmer. According to the recent shift in industrial trends, including the high-mix low-volume production, the robots need to have a capability of the flexibility to adapt dynamic
changes in the environment and an applicability in fields out of factories for interacting with human workers, like
agricultural purposes. Description logics (DLs) was used in automated planning and scheduling for robots and Planning Domain Definition Language (PDDL) is a typical method for the actual implementation into the robot with
real-time computation for planning. In parallel, semantic web technologies for industrial engineering are recently
highlighted for needs of the expert knowledge transfer and risk assessments in manufacturing companies. Web
Ontology Language (OWL) is a description standardized by World Wide Web Consortium (W3C). For the high-level
flexibility in the dynamic environment, there are approaches based on Linear Temporal Logic (LTL) for conflict
managements with other entities and a modified version of OWL-DL reasoner to execute Horn clauses yet regulate
the reasoning range to prevent the combinatorial explosion of possibilities in the world, known as task ontology,
which is a practical solution to resolve the frame problem. In this seminar, the importance of DLs for automated
planning and scheduling for robots is discussed in the context of industry needs. It will be an opportunity to explore
possible mechanisms of the reasoner to execute Horn clauses in complex circumstances toward competencies in
human expertise.