AI Planning and Machine Intelligence
Grupo de Pesquisa

AI Planning and Machine Intelligence group studies classic and of knowledge engineering to support intelligent planning applied to automation. The assumption is that intelligent automation could be based on two different paradigms: knowledge engineering and requirements modeling and analysis, and the synthesis of a planning domain, composed by a planning problem and the model of a plan environment - where the plan is occurring. Specially we work in a comparison between the requirements acquisition using different modeling languages and its further modeling and analysis in Petri Nets to turn out in formal specifications that could be delivered to existing plans which use different planning approaches. Special attention is given to Graphplan and to JStore - to explore hierarchical approach.
Based o this concepts we propose an advanced approach directed to behavior (instead of just actions) which potentialize the cognitive plan applied to robots and complex automated systems. That could also be applied to critical systems, where failure is not admitted.
Project associated with research group is itSIMPLE (Integrated Tools and Software Interface for Modeling Planning Environment), which is not in its version 4.0. A new version 5.0 is under development with new features and a new approach to requirements analysis for planning and scheduling. A new package to perform the analysis of cognitive behavior in planning will be also included.
Atualizado em
03 de Julho de 2024, 22:44
Pesquisador Responsável
Jose Reinaldo Silva
reinaldo@usp.brBachelor in Physics from Bahia Federal University, got aMSc in Physics from Pernanbuco Federal University UFPE, and a MA in Computer Science from Mills College, USA. PhD in Computer Engineering by Universith of São Paulo and postdoc in Computer Science and in Systems Design Engineering at University of Waterloo, Ca. Research interests are in Engineering Systems Design, Requirements Engineering, Service Design Management and Engineering, Artificial Intelligence, Planning & Scheduling, Formal Verification methods, Data Spaces and Digital Transformation towards Industry 4.0.
Colaboradores
Elinilson Vital
vital@usp.brMaster's student in Mechatronics Engineering at the Polytechnic School of the University of São Paulo, focusing on the theme Cloud Manufacturing Services (CMfgS): Modeling of automated production systems in the cloud. Graduated in Applied Mathematics in Mechanical Control and Automation from the Institute of Mathematics and Statistics of the same university, focusing on formal requirements analysis, service science, distributed systems, knowledge bases, and data spaces. Demonstrates ability to apply mathematical and computational knowledge to solve complex problems and promote technological innovations. Practical experience includes data science, data grouping and their applications, as well as in modeling and analysis of automation systems, also in programming and simulation, including algorithms and data structures, concurrent and parallel programming, and numerical methods for differential equations. The main research interests are currently focused on deepening knowledge in Service Science and Data Spaces, aiming to develop innovative solutions that optimize business processes and increase technological efficiency in distributed environments, uniting mathematics, computing, and engineering.
Javier Martinez Silva
javier.cu@gmail.comGraduated in Computer Science from Universidad de Oriente, Cuba and received a Master in Computer Science from this same institution. Ph.D. in Mechatronics (Mechanical Control and Automation) from Escola Politécnica, USP, working with D-Lab. Postdoc in D-Lab, where he also acted on implementing the USP-AWS agreement. His research interests are in formal methods for modeling and analysis of requirements in automation projects and working with AI Planning to design modeling. Worked in the Samsung Research Center and is now working in Instituto Eldorado, another research institution linking the academy and the market.
Luiz Fernando Ferreira da Silva
luizffs@usp.brGraduated in Mechatronics by Escola Politécnica da USP with academic interests in programming and requirements analysis. Had an internship in MVISIA and work today ag WEG developing computer vision solutions in Python. Started to work in D-ab directing his interests to obtain a Master degree in the area of AI Planning.
Matheus Alexandrino Brito
matheus.a.brito12@usp.brMatheus is a undergad student in the Mechatronics course and has a scholarship from USP, associated to D-Lab. His interests are in automation processes, particularly those involving Artificial Intelligence and AI Planning. Therefore, he join the research line in AI Planning and Machine Intelligence and do the project DiPlant (Didatic Planning Tool) - a system for analyzing plans (or post-planning) expected to be used in the undergrad course PMR3510 Inteligência Artificial.
Yaney Gomez Correa
ygc8104@usp.brGraduate in Automatic Control at Universidade de Oriente, Cuba, and worked in automated systems associated to power supply and petroleum in this country. Got a MSc. at D-Lab in the subject of Service Automation Design applied to Healthcare, proposing assistant services on top of a building and residence automation. Ph.D. student at D-Lab working in Model-based Requirements and Hierarchical AI Planning.
Projetos

D4PlanS - Design for Planning Services
D4PlanS é uma rede de micro serviços integrada em um ambiente de nuvem para o design de sistemas de planejamento baseados em Inteligência Artificial. O ambiente de design usa técnicas de design de planos integradas ao métodos de design de sistemas orientados a objetivos, e formalizados em Redes de Petri. HTN (Hierarchical Task Network) é a base da modelagem, e o conceito de "task" é reinterpretado como serviço, compondo uma rede generativa para um plano de nível mais alto. O...