Printfarm Management System
Author
Aagren, Alexander
Term
4. Semester
Education
Publication year
2023
Submitted on
2023-06-02
Pages
58
Abstract
Dette speciale er udarbejdet i samarbejde med 3D‑printfirmaet Create it REAL. Det undersøger, hvordan man bedst driver en 3D‑printerfarm (en samling af mange 3D‑printere), både set fra hardware‑siden og i forhold til netværksforbindelser. Undersøgelsen viser, at printjobs sendt over Wi‑Fi til printere med Create it REALs hardware og firmware er beskyttet med AES‑128‑kryptering, en udbredt standard til at sikre data. Der er designet et Printfarm Management System, som gør det muligt at lægge flere printjobs i kø til hver printer. Systemet planlægger efter, hvor lang tid en opgave forventes at tage, så det kan vurdere, om printet kan blive færdigt inden for arbejdstiden. Det tager også højde for den anslåede mængde filament, der er tilbage på den enkelte printer, så administratoren kan få besked, hvis der er behov for at skifte filament. Flere køalgoritmer blev undersøgt og simuleret for at finde den bedst egnede løsning. Valget faldt på Modified Maximum Urgency First (en planlægningsmetode, der kan prioritere presserende opgaver), fordi den giver mulighed for at fremhæve særligt vigtige printjobs og dermed understøtte håndtering af fejlramte print.
This thesis was carried out in collaboration with the 3D printing company Create it REAL. It examines how to run a 3D printer farm (many printers managed together), looking at both hardware and network connectivity. The study found that print jobs sent over Wi‑Fi to printers using Create it REAL hardware and firmware are protected with AES‑128 encryption, a widely used standard for securing data. A Printfarm Management System was designed to queue multiple print jobs per printer. The system schedules jobs by considering the expected printing time, so it can determine whether a print will finish within working hours. It also tracks the estimated filament remaining on each printer to notify the administrator when a filament change may be needed. Several queuing algorithms were investigated and simulated to select the most suitable one. The chosen approach was Modified Maximum Urgency First (a scheduling method that prioritizes urgent tasks), because it allows specific print jobs to be prioritized and supports handling of failed prints.
[This abstract was generated with the help of AI]
Keywords
3D Printing ; Algorithms ; Print farm ; Printfarm ; Queuing
Documents