International Journal of Electrical and Computer Engineering (IJECE) Vol. 12, No. 3, June 2022, pp. 2425~2431 ISSN: 2088-8708, DOI: 10.11591/ijece.v12i3.pp2425-2431  2425 Journal homepage: http://ijece.iaescore.com Robot for plastic garbage recognition Janusz Bobulski, Mariusz Kubanek Department of Computer Science, The Czestochowa University of Technology, Czestochowa, Poland Article Info ABSTRACT Article history: Received Jun 1, 2021 Revised Dec 23, 2021 Accepted Jan 5, 2022 Waste and related threats are becoming more and more severe problems in environmental security. There is growing attention in waste management globally, both in developing techniques to decrease their quantity and those correlated to their neutralization and commercial use. The basic segregation process of waste due to the type of material is insufficient, as we can reuse only some kinds of plastic. There are difficulties with the effective separation of the different kinds of plastic; therefore, we should develop modern techniques for sorting the plastic fraction. One option is to use deep learning and a convolutional neural network (CNN). The main problem that we considered in this article is creating a method for automatically segregating plastic waste into seven specific subcategories based on the camera image. The technique can be applied to the mobile robot for gathering waste. It would be helpful at the terrain and the sorting plants. The paper presents a 15-layer convolutional neural network capable of recognizing seven plastic materials with good efficiency. Keywords: Artificial intelligence Deep learning convolutional neural network Environment protection Image processing Waste management This is an open access article under the CC BY-SA license. Corresponding Author: Janusz Bobulski Department of Computer Science, The Czestochowa University of Technology 73 Dabrowskiego Str., 42-201 Czestochowa, Poland Email: januszb@icis.pcz.pl 1. INTRODUCTION An intensive increase in the consumption of main raw materials causes an increase in waste appropriate for reuse, which absurdly reduces resource use. Due to the constantly growing raw material needs in the industry and construction sector, it is important to gradually increase waste as secondary raw materials as high as possible. The research results presented that the expenses of obtaining raw materials are greater than gathering and regenerating secondary raw supplies acquired from manufacture or post-use waste. Gathering and recyclable old substances also involve minor energy use than producing a new one [1]. We may use municipal and agricultural waste for the production of gas or heat energy. Changing original materials with processed materials also reduce the use of resources, excludes the cost of transferring waste to landfills, maintains, reduces work input, and decreases the production costs of products. Regulating the mass of produced waste to a quantity ensuring raw material, biological and hygienic balance is impossible without far-reaching harmonisation of knowledge and the way people live with the development and functioning of an ecological arrangement in a given region. Activities goal at decreasing the quantity of waste formed and collected in the environment contain reprocessing raw materials, reducing waste creation through up-to-date low-waste or non-waste technologies, and exchanging used materials [2]. The aim of society for solving fabrication waste polluting is minor and waste-free knowhow. Non-waste technology (NWT) is created with avoiding waste and complete use of substances. It implicates several technical procedures that lead to total supervision and the removal of pollution without damaging effects on the environs. The primary disorder has not deposited the waste. The application of NWT has monetary motivation for the reason of the complete use of materials. Reducing the number of waste permits