Keynote Speakers:

Lecture topic: Artificial Intelligence: Concerns and Considerations

Presenter: Prof. Mohammad Bagher Menhaj (Amirkabir University of Technology, Tehran, Iran)

Talk orientation: electricity and computer

Time: 30 minutes

Abstract:

In this lecture, we will discuss the evolution and development of AI. It has received much attention, especially in the last two decades. In all technical and engineering fields and specifically in decision-making and control systems, AI techniques are used for data processing and distributed implementation of decision-making and control systems. For example, in the feedforward path, classical systems can be improved by predicting the output of the decision-making system. Then, these predictions are used for error screening. One of the latest strategic developments in this period is decision engineering inspired by cognitive-control processes based on reinforcement learning and cognitive control. Here, the executive functions of cognitive processes are investigated in two permanent and changing environments.

In this meeting, by covering classic cognitive research and theories as well as neuroscience, we will have a deep look at these processes from the perspective of systems theory. Identification of systems, cognitive robotics, design and construction of machine-brain interfaces, development of cognitive learning systems, computational neuroscience, and analysis of the relationship between functional structural features of the brain and their dynamic changes in different time scales will be considered in this meeting. Applied computational intelligence tools in decision-making engineering include knowledge-based systems, fuzzy calculations, neural networks, and evolutionary algorithms. The old AI technologies could not store information and make decisions based on past experiences because they could not make inferences and interpolate. In other words, the previous systems could not approximate the functions from the data. With the increase in the speed of information processing, new AI algorithms were developed for data analysis.

Currently, with the advent of cloud and distributed computing, the analysis of complex systems is done in the general framework of cognitive computing. However, as the data volume increases, there are still more computational requirements. Since AI systems do not get tired, frustrated, and distracted, unlike humans, if they are equipped with robust algorithms and software, they will be able to process much more information than we can imagine. Therefore, it seems that in the future, AI will be responsible for creating everything from computer code to visual arts. This will lead to the dimming of the role of skill and creativity in humans. Other AI problems include job displacement, ethical concerns about privacy, and security risks caused by hacking. In this meeting, we will have more discussion on AI issues and we will think and meditate to overcome these threats.

-------------------------------------------------------------------------------------------

Lecture topic: Decision Support and Intelligent Systems in Container Terminals

Presenter: Prof. Hassan Rashidi (Allameh Tabataba'i University (ATU))

Talk orientation: electricity and computer

Time: 1 hour

Abstract:

This talk presents the optimization problems faced by today’s modern container terminals, in general, and then focuses on the scheduling problem of automated guided vehicles (AGV), in particular. This particular problem is defined and formulated as a Minimum Cost Flow model. This talk presents a novel algorithm, NSA+, which extends the standard Network Simplex Algorithm (NSA) for the problem. To complement NSA+, an incomplete algorithm Greedy Vehicle Search (GVS) is presented for the problem.  The NSA+ and GVS are compared and contrasted to evaluate their relative strength and weaknesses. With polynomial time complexity, NSA+ can be used to solve very large problems, as verified in the experiments. Should the problem be too large for NSA+, or the time available for computation is too short (as it would be in dynamic scheduling), GVS complements NSA+. The most important topics of this speech are:

• Challenges in Container Terminals
• Components in the container terminals
• Classification of Scheduling problems in Ports
• The Problem of AGV Scheduling and its modelling
• Algorithms for the Problem of AGV Scheduling