DigitaLLeather
PROBLEM WE ARE SOLVING
70-85%
Accuracy
Operator accuracy directly affects the utilization of industrial leather
50-85%
Usage
Material utilization is key in the leather industry
ROADMAP
Founding of DigitaLLeather
Start-up DigitaLLeather was founded within the early development program of the Fund for Innovation Activities of the Republic of Serbia.
Proof of concept (Prototype P1-22)
The project lasted one year and the outcome of the project was the developed prototype P1-22, which proved the concept of a solution for quality control of industrial leather using artificial intelligence. The P1-22 prototype achieved an acceptable degree of accuracy in detecting defects on industrial leather. The prototype is at the TRL4 level of development and has been tested in laboratory conditions.
Prototype P2-24
After successfully participating in the acceleration program of the United Nations Development Program (UNDP), we received a grant under the Green Agenda program for Serbia to develop the improved Prototype P2-24, which is at the TRL6 level of development. It is currently in the testing phase, and we expect the first results at the end of 2024.
Scaled solution (Prototype FHP1-25)
The Fund for Innovation Activity supported a scaled solution for quality control of entire industrial hides using artificial intelligence. Unlike previous small-scale solutions, this solution will contain a vacuum table and a chamber 4 times larger than the P2-24 prototype that can fit an entire industrial skin. The solution is at the technical level of TRL5 development.
Market breakthrough
We have been doing the entire research since the beginning with the company AutoStop Interiors doo from Leskovac, which is part of AutoStop Global Ltd. groups. AutoStop interiors supplies us with defective industrial leather and the idea is to make a market breakthrough precisely by adopting our solutions by this company.
INSTITUTIONS THAT SUPPORTED US
Naša Rešenja
P1-22
Prototip razvijen u okviru programa ranog razvoja, 2021-2022 godine. Ovim prototipom je dokazan koncept rešenja i mogućnost kontrole kvaliteta primenom veštačke inteligencije. Prosečna postignuta tačnost je bila 74%, dok tačnost operatera kontrole kvaliteta varira između 70-85%. Rešenje je testirano u kompaniji AutoStop Interiors na primerima kvalitetnih i defektnih isečaka (Cut parts kad budeš prevodio) industrijske kože.
P2-24
Ovaj prototip je podržan od strane Razvojnog Fonda Ujedinjenih nacija i ima za cilj obostranu digitalizaciju industrijske kože, da bi naličje služili kao podrška licu pri otkrivanju, a u cilju povećanja stepena tačnosti. Prototip je namenjen kontroli kvaliteta isečaka industrijske kože primenom veštačke inteligencije. Prednost ovog prototipa je kontrola dimenzija defekata čime se može proveriti da li su defekti u toleranciji.
FHP1-25
Ovaj prototip je skalirano rešenje prethodna dva prototipa (P1-22 i P2-24) namenjena za otkrivanje defekata na isečenim delovima industrijske kože. Ovo rešenje je 2x2m dimenzija i namenjeno je kontroli kvaliteta celih industirjskijh koža primenom veštačke inteligencije. Hardver je znatno složeniji jer sadrži vakum sto i senzore kao pomoć pri pozicioniranju i ravnanju industrijske kože. Trenutno je u fazi razvoja.
Tim
Naš tim
Nikola Banduka, Doktor nauka iz oblasti industrijskog
inženjerstva
CEO
Katarina Tomić, Master inženjer informacionih tehnologija
Inženjer kompjuterske vizije
Nikola Džajević, Diplomirani inženjer elektrotehnike i računarstva
Glavni Backend developer
Milovan Marjanović, Diplomirani inženjer informacionih tehnologija
Glavni Frontend developer
Milan Ivković, Master inženjer
mašinstva
Mašinski inženjer
Kristina Mojsić, Master inženjer elektrotehnike i računarstva
Backend developer
Jovan Živadinović, Master inženjer elektrotehnike i računarstva
Backend developer
Pavle Rvović, Master urbanog
inženjerstva
Frontend developer
Lazar Milošević, Master inženjer informacionih tehnologija
Inženjer elektrotehnike
