Upravljanje robotom temeljeno na video informaciji na studijskom slučaju igranja šaha

master's thesis

• Title (english): Robot control based on video information on a chess study case study
• Author: Domagoj Kudek
• Mentor: Mirko Sužnjević (mentor)
• Committee member: Mirko Sužnjević (predsjednik povjerenstva)
• Committee member: Maja Matijašević (član povjerenstva)
• Committee member: Lea Skorin-Kapov (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2021-07-12, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Automatizacija industrijskih pogona i uporaba umjetne inteligencije u robotici jedna je od ključnih točaka industrije 4.0. Sustavi koje se trenutno koriste diljem svijeta uglavnom su rađeni namjenski, što znači da je njihova ponovna uporaba za druge svrhe komplicirana i vremenski dugotrajna. Kako bi napravili univerzalna rješenja, koja bila omogućila jednostavniji razvoj, moramo razviti sustav temeljen na poznatim programskim jezicima i platformama. Ocellus, kojeg je razvila švedska tvrtka Byte Motion, je jedan primjer takvog sustava. Razvijen je u poznatom sustavu za razvoj igara pod nazivom Unity i podržava širok spektar hardverskih i softverskih komponenti. U ovom diplomskom radu, koristit ćemo Ocellus i njegove napredne module pogonjene umjetnom inteligencijom kako bi razvili sustav koji ćemo nazvati AutoCHESS. AutoCHESS je potpuno automatizirani sustav koji simulira igru partiju šaha između dvije robotske ruke. Uz pomoć dubinske kamere RealSense, neuronska mreža trenirana za otkrivanje šahovskih figura, slat će informacije Ocellusu koji će računati njihove stvarne položaje. Informacije će se prenositi putem Node-REDA, alata za vizualno programiranje koji se inače koristi za uređaje interneta stvari. Potezi šahovske igre također će biti automatizirani i izračunati putem trenutno najboljeg šahovskog kalkulatora poteza pod nazivom Stockfish. Generirane poruke RMQ protokola slat će se s Node-REDA na ABB-ov program RobotStudio na obradu. RobotStudio će tada izvršiti radnje koje će rezultirati kretanjem robotskih ruku koje će naizmjenično odigravati šahovske poteze sve dok jedan od igrača ne pobijedi. Gotov sustav poslužit će kao demonstracija mogućnosti upravljačkih sustava za robote temeljenih na sustavima za razvoj igara, kao i prikaz upotrebe naprednih algoritama umjetne inteligencije u industrijskoj robotici.

Abstract (english)

Industrial automation and the use of artificial intelligence in robotics is one of the key focus points of Industry 4.0. Currently, systems used around the world are mostly purpose-built, which means that their reuse in other applications is difficult and time-consuming. To make universal solutions, that are easier to develop, we must construct a system out of well-known programming languages and platforms. Ocellus, developed by a Swedish firm called Byte Motion, is one such system. It is developed in a game engine called Unity and it supports a wide range of hardware and software components. In this thesis, we are going to use Ocellus and its advanced artificial intelligence modules to create a system shorty called AutoCHESS. AutoCHESS is a fully automated system that simulates a game of chess played by two robot arms against each other. With the help of a RealSense depth camera, a neural network trained to detected chess figures will supply information to Ocellus that will calculate their real-world positions. This information will be transferred via Node-RED, a visual programming tool that is mostly used on IoT devices. The moves of the chess game will also be automated and calculated by the currently best open-source chess engine called Stockfish. Constructed RMQ protocol messages will be sent from Node-RED to ABB’s RobotStudio for processing. RobotStudio will then execute procedures that will result in movement of the robot arms that will intermittently complete chess moves until one of the players wins the game. The finished system will serve as a demonstration of the capabilities of game engine-based robot control systems as well as a showcase of the use of advanced artificial intelligence algorithms in industrial robotics.

• Keywords: Unity; Ocellus; kontroler robota; umjetna inteligencija; neuronske mreže; strojni vid; Node-RED; ABB RobotStudio; industrijska automatizacija
• Keywords (english): Unity; Ocellus; robot controller; artificial intelligence; neural networks; machine vision; Node-RED; ABB RobotStudio; industrial automation
• Language: english
• URN:NBN: urn:nbn:hr:168:396314
• Study programme: Title: Computing; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-01-13 12:49:32