multi-agent.io

Laboratory of Multi-Agent Systems in Smart Cities and Industry 4.0

Lab Logo

The laboratory specializes in the study of issues of cybernetics and economic theory as applied to communication in multi-agent systems.

Examples of issues to be considered are:

The priority tasks are:

Lab tasks: https://github.com/orgs/airalab/projects/9

Program

Name

Laboratory of Multi-Agent Systems in Smart Cities and Industry 4.0

MASSCI4.0

Brief Description

Laboratory for the study of aspects of the organization of networked cyber-physical systems in autonomous conditions using decentralized technologies

Team experience

Vadim Manaenko - head of the laboratory, ICT assistant, postgraduate student of the 2nd year of study.

Recent projects:

Ivan Berman is a scientific administrator, an engineer of CSR, a post-graduate student of the 2nd year of study.

Recent projects:

Alexander Kapitonov - curator of the laboratory, Ph.D., dean of ICT.

Keywords

Research area and objectives

The ongoing large-scale digitalization of human activity has been made possible by the common network infrastructure offered by the World Wide Web, as well as numerous devices, from CCTV systems and sensors to smartphones and smart devices.

Given the growing influence of robotics, a logical step should be the same digitalization of cyber-physical systems and devices themselves, i.e., their large-scale network integration to interact with human infrastructure and with each other. This will allow devices to use the collected data to self-manage and provide the most effective services to the person. Researchers have called this approach Robot-as-a-Service and see it as a potential economic model for the future.

However, the multitude of autonomous devices that are networked and capable of physically affecting the world raise great security concerns, exacerbated by the negative experience of large centralized projects. Such network CPSs require studying the issues of communication protocols and managing multi-agent systems of heterogeneous devices, without taking into account the fact that managing a single device can also be a nontrivial task. In addition, the final goal of creating a networked CPS is its integration into the digital economy, and this imposes additional restrictions related to legal and economic aspects and information protection.

Decentralized technologies such as peer-to-peer networks, blockchains and smart contracts are a promising area of research in this area. They do not have the drawbacks of centralized approaches, make it easier to solve some issues of organizing a multi-agent system, and also already have ways of integrating with the digital economy.

Therefore, within the framework of this laboratory, we propose to investigate aspects of the application of precisely decentralized technologies for managing network CPS. It is supposed to use the open source platform Robonomics, which offers the capabilities of the Ethereum and Polkadot blockchains for robotic devices compatible with the Robot Operating System framework, and uses the market mechanism as the basic way of interaction between devices and humans. The main goal of the laboratory is the application of decentralized technologies for networked cyber-physical systems in various applications.

Relevance to ITMO’s prioritized areas

The research interest of the laboratory corresponds to the priority areas of ITMO University in “Intelligent technologies and robotics”, as well as on “Intelligent technologies in the economy, social sphere and art”.

Research plan

The laboratory focuses on four major areas of research:

1) Certified measurements of the quality of water resources using unmanned surface vehicles.

The main focus is on an environmental monitoring system built on a small fleet of small marine vessels equipped with compact sensors. In this case, the Robnonomics platform is used for secure registration of vessels, their missions and collected data on water samples, thereby ensuring the certification of measurements. Within the framework of the direction, the following issues are investigated:

2) Smart factories for Industry 4.0.

Smart factories are factories in which autonomous production occurs by collecting, exchanging and using information between people, materials and machines using network technologies. In the final stage, such factories operate without human intervention, independently adjusting their production process based on feedback from product sales. A test bench is available at the faculty, based on the Fischertechnik designer and Siemens controllers, which simulates a smart factory, from the supply of raw materials to the shipment of products from warehouses, depending on the demand for it. The stand allows you to explore the following issues:

3) The robot artist as an autonomous creative system.

Autonomous systems that create art objects (or artifacts) have been of interest to researchers since the 1980s, gradually accelerating with the development of image recognition and Simultaneous Localization and Mapping (SLAM) techniques. However, the task of generating a human-independent idea of an artifact - what is called creativity - has begun to advance recently, with the development of neural networks, for example, art in the style of Inceptionism from the Google AI team. The most intriguing thing is the creation of a completely autonomous creative system capable of “inventing” the idea of an artifact, realizing it, presenting it for review, and drawing conclusions from feedback on how to proceed further. The KUKA KR6 R900 manipulator is available to work on this direction at the faculty, which can draw with a brush on canvas. The robot works on the basis of the Robonomics platform, which allows it, for example, to put up paintings for a digital auction. The stand allows you to explore the following issues:

4) Digitalization of the state of the infrastructure using a network of sensors.

The Internet of Things has been actively expanding in the last decade, but it still has significant limitations for ubiquitous use, primarily related to security when connecting smart devices and sensors. Decentralization of governance has the potential to circumvent this limitation, opening the way for decentralized autonomous organizations and even smart cities. This infrastructure has a digital footprint for every facility in the organization, allowing you to analyze health and efficiently manage assets and costs, from electricity consumption to PM10 concentration in offices. Within the framework of this direction, it is planned to investigate:

Skills that participants will develop

First year plan

Materials

Please find the Russian version here