Publications Repository - Gdańsk University of Technology

Page settings

polski
Publications Repository
Gdańsk University of Technology

Show publications from the year

Publications from the year 2024

Show all
  • Advancing sustainable hybrid bitumen systems: A compatibilization solution by functionalized polyolefins for enhanced crumb rubber content in bitumen
    • Mateusz Malus
    • Joanna Bojda
    • Maciej Sienkiewicz
    • Wojciech Szot
    • Miloud Bouyahyi
    • Lanti Yang
    • Francisco Javier Navarro
    • Maha AlSayegh
    • Rasha Daadoush
    • Maria Soliman
    • Rob Duchateau
    • Lidia Jasinska-Walc
    2024 JOURNAL OF CLEANER PRODUCTION

    Polymer waste pollution has a profound effect on the environment and, consequently, on the lifestyle of hu- mankind. The massive production and disposal of cross-linked polymers clearly exemplify the challenges of recycling. Increasing efforts are being undertaken to introduce recycled polymers, especially crumb rubber (CR), into asphalt formulations. Due to the rather poor processability and phase separation associated with CR- modified bitumen (CRMB) compositions, a broader implementation of such concept is challenging unless an efficient compatibilizer is applied. Results from the study on usage of In-Reactor-Functionalized Polyolefins viz. poly(propylene-co-hex-1-ene-co-hex-5-en-1-ol) (FPP), demonstrated excellent compatibilizing ability in CRMB, allowing incorporation of up to 10 wt% of CR. This represents a significant improvement when compared to the best-in-class solutions. The FPP-containing products exhibit superior bulk, nanomechanical and rheological properties, as well as stability during binder annealing. Furthermore, the bitumen surface morphology is significantly improved. The polar groups present in the FPP create a thermo-reversible interpenetrating cross- linked network that provides mechanical integrity and contributes to the adhesion to different components of the modified bitumen at service temperatures, enhancing its processability. The exceptional compatibility of FPP in CRMB resulted in a significant increase in the Performance Grade of the hybrid system by 5 classes (88) compared to neat bitumen (58). Moreover, the best-performing composition fulfilled the low-temperature ductility specifications, withstanding deformation without fracturing or breaking up to a 400 mm elongation.

  • Advancing sustainable wastewater management: A comprehensive review of nutrient recovery products and their applications
    • Bogna Śniatała
    • Hussein Al-Hazmi
    • Dominika Sobotka
    • Jun Zhai
    • Jacek Mąkinia
    2024 SCIENCE OF THE TOTAL ENVIRONMENT

    Wastewater serves as a vital resource for sustainable fertilizer production, particularly in the recovery of nitrogen (N) and phosphorus (P). This comprehensive study explores the recovery chain, from technology to final product reuse. Biomass growth is the most cost-effective method, valorizing up to 95 % of nutrients, although facing safety concerns. Various techniques enable the recovery of 100 % P and up to 99 % N, but challenges arise during the final product crystallization due to the high solubility of ammonium salts. Among these techniques, chemical precipitation and ammonia stripping/ absorption have achieved full commercialization, with estimated recovery costs of 6.0–10.0 EUR kgP-1 and 4.4-4.8 £ kgN-1, respectively. Multiple technologies integrating biomass thermo-chemical processing and P and/or N have also reached technology readiness level TRL = 9. However, due to maturing regulatory of waste-derived products, not all of their products are commercially available. The non-homogenous nature of wastewater introduces impurities into nutrient recovery products. While calcium and iron impurities may impact product bioavailability, some full-scale P recovery technologies deliver products containing this admixture. Recovered mineral nutrient forms have shown up to 60 % higher yield biomass growth compared to synthetic fertilizers. Life cycle assessment studies confirm the positive environmental outcomes of nutrient recycling from wastewater to agricultural applications. Integration of novel technologies may increase wastewater treatment costs by a few percent, but this can be offset through renewable energy utilization and the sale of recovered products. Moreover, simultaneous nutrient recovery and energy production via bio-electrochemical processes contributes to carbon neutrality achieving. Interdisciplinary cooperation is essential to offset both energy and chemicals inputs, increase their cos-efficiency and optimize technologies and understand the nutrient release patterns of wastewater-derived products on various crops. Addressing non-technological factors, such as legal and financial support, infrastructure redesign, and market-readiness, is crucial for successfully implementation and securing the global food production.

  • Advancing Urban Transit: Gepard and CAR projects - Innovations in Trolleybus Technology
    • Mikołaj Bartłomiejczyk
    • Leszek Jarzębowicz
    • Slobodan Mirchevski
    • Marcin Połom
    2024

    The Gepard project in Gdynia, Poland, revolutionized the city's trolleybus network with the introduction of “Trolleybus 2.0” vehicles and an innovative charging system. “Trolleybus 2.0” vehicles combine features of traditional trolleybuses and electric buses boasting traction batteries for autonomous driving and dual legal approval. Statistical analysis of energy consumption informed the development of a hybrid charging concept, balancing overhead contact line (OHL) coverage with additional fast charging stations. This hybrid In Motion Charhing (IMC) system reduces costs while ensuring reliable operation, even in adverse weather conditions. Moreover, as part of the CAR project, a fast charging station for trolleybuses was constructed, allowing for the additional extension of trolleybus routes.

  • A Flexible Way of Coarse Coordinates Estimation for Sodars
    • Kamil Stawiarski
    2024 International Journal of Electronics and Telecommunications

    The publication presents a flexible approach to implementing coarse coordinate estimation of an object observed with a sodar. This flexibility permits any arrangement of sound sources as well as microphones. Only minimal requirements are imposed on the probing signal, which can particularly be broadband. The algorithms have been tested on both synthetic data and data recorded with an actual device.

  • AGREEMIP: The Analytical Greenness Assessment Tool for Molecularly Imprinted Polymers Synthesis
    • Mariusz Marć
    • Wojciech Wojnowski
    • Francisco Pena-Pereira
    • Marek Tobiszewski
    • Antonio Martín-Esteban
    2024 ACS Sustainable Chemistry & Engineering

    Molecular imprinting technology is well established in areas where a high selectivity is required, such as catalysis, sensing, and separations/sample preparation. However, according to the Principles of Green Chemistry, it is evident that the various steps required to obtain molecularly imprinted polymers (MIPs) are far from ideal. In this regard, greener alternatives to the synthesis of MIPs have been proposed in recent years. However, although it is intuitively possible to design new green MIPs, it would be desirable to have a quantitative measure of the environmental impact of the changes introduced for their synthesis. In this regard, this work proposes, for the first time, a metric tool and software (termed AGREEMIP) to assess and compare the greenness of MIP synthesis procedures. AGREEMIP is based on 12 assessment criteria that correspond to the greenness of different reaction mixture constituents, energy requirements, and the details of MIP synthesis procedures. The input data of the 12 criteria are transformed into individual scores on a 0−1 scale that in turn produce an overall score through the calculation of the weighted average. The assessment can be performed using user-friendly open-source software, freely downloadable from mostwiedzy.pl/agreemip. The assessment result is an easily interpretable pictogram and visually appealing, showing the performance in each of the criteria, the criteria weights, and overall performance in terms of greenness. The application of AGREEMIP is presented with selected case studies that show good discrimination power in the greenness assessment of MIP synthesis pathways.

  • Agri-food waste biosorbents for volatile organic compounds removal from air and industrial gases – A review
    • Patrycja Makoś-Chełstowska
    • Edyta Słupek
    • Jacek Gębicki
    2024 Full text SCIENCE OF THE TOTAL ENVIRONMENT

    Approximately 1.3 billion metric tons of agricultural and food waste is produced annually, highlighting the need for appropriate processing and management strategies. This paper provides an exhaustive overview of the utilization of agri-food waste as a biosorbents for the elimination of volatile organic compounds (VOCs) from gaseous streams. The review paper underscores the critical role of waste management in the context of a circular economy, wherein waste is not viewed as a final product, but rather as a valuable resource for innovative processes. This perspective is consistent with the principles of resource efficiency and sustainability. Various types of waste have been described as effective biosorbents, and methods for biosorbents preparation have been discussed, including thermal treatment, surface activation, and doping with nitrogen, phosphorus, and sulfur atoms. This review further investigates the applications of these biosorbents in adsorbing VOCs from gaseous streams and elucidates the primary mechanisms governing the adsorption process. Additionally, this study sheds light on methods of biosorbents regeneration, which is a key aspect of practical applications. The paper concludes with a critical commentary and discussion of future perspectives in this field, emphasizing the need for more research and innovation in waste management to fully realize the potential of a circular economy. This review serves as a valuable resource for researchers and practitioners interested in the potential use of agri-food waste biosorbents for VOCs removal, marking a significant first step toward considering these aspects together.

  • AI-Powered Cleaning Robot: A Sustainable Approach to Waste Management
    • Johan Carcamo Pineda
    • Ahmad M.A. Shehada
    • Arda Candas
    • Nirav Vaghasiya
    • Murad Abdullayev
    • Jacek Rumiński
    2024

    The world is producing a massive amount of single use waste, especially plastic waste made from polymers. Such waste is usually distributed in large areas within cities, near roads, parks, forests, etc. It is a challenge to collect them efficiently. In this work, we propose a Cleaning Robot as an autonomous vehicle for waste collection, utilizing the Nvidia Jetson Nano platform for precise arm movements guided by computer vision capabilities. Integrated with the Raspberry Pi platform for mobility control, the robot employs the YOLO (You Only Look Once) framework for efficient waste detection and classification. The model was trained, implemented in the robot prototype, and tested on preprocessed waste images, resulting in mean average precision (mAP) above 80 percent. Our design emphasizes singular-object focus, enabling real-time detection of waste with accurate distance (83.7-95.6%) and direction (84.7 97.3%) information. The robot autonomously navigates towards detected waste, halting at a predefined distance for collection and disposal into a designated bin. This work contributes to advancements in waste management systems using small robots.

  • AI-powered Customer Relationship Management – GenerativeAI-based CRM – Einstein GPT, Sugar CRM, and MS Dynamics 365
    • Edyta Gołąb-Andrzejak
    2024 Full text Procedia Computer Science

    Generative artificial intelligence (GenAI) and its implementation in successive business management support systems is a rapidly growing area of theoretical consideration, ongoing research, discourse and application in practice. Recently, the implementation of of GenAI in customer relationship management (CRM) systems has been observed. Accordingly, the aim of this article is to identify areas where GenAI can enhance CRM systems, using Einstain GPT, Sugar CRM or Microsoft Dynamics 365 as examples. To this end, a research question was formulated: how can GenAI improve the effective use of CRM systems? Accordingly, a preliminary study based on secondary data analysis as well as software analysis was conducted to identify areas of GenAI use in CRM systems where we see an increase in the effective application of CRM. The results of the analysis showed that GenAI-powered CRM systems support the effectiveness and efficiency of marketing, sales, commerce, service and system user success. This is because they provide numerous advantages in terms of developing, expanding and strengthening customer relationships through highly advanced personalisation, closely linked to customer segmentation, which allows unique experiences to be provided to individual segments. As a result, this translates into building a company's competitive advantage and increasing the profitability of its CRM efforts.

  • Airport Runoff Water: State-of-the-Art and Future Perspectives
    • Anna Maria Sulej-Suchomska
    • Danuta Szumińska
    • Miguel de la Guardia
    • Piotr Przybyłowski
    • Żaneta Polkowska
    2024 Sustainability

    The increase in the quantity and variety of contaminants generated during routine airport infrastructure maintenance operations leads to a wider range of pollutants entering soil and surface waters through runoff, causing soil erosion and groundwater pollution. A significant developmental challenge is ensuring that airport infrastructure meets high-quality environmental management standards. It is crucial to have effective tools for monitoring and managing the volume and quality of stormwater produced within airports and nearby coastal areas. It is necessary to develop methodologies for determining a wide range of contaminants in airport stormwater samples and assessing their toxicity to improve the accuracy of environmental status assessments. This manuscript aims to showcase the latest advancements (2010–2024 update) in developing methodologies, including green analytical techniques, for detecting a wide range of pollutants in airport runoff waters and directly assessing the toxicity levels of airport stormwater effluent. An integrated chemical and ecotoxicological approach to assessing environmental pollution in airport areas can lead to precise environmental risk assessments and well-informed management decisions for sustainable airport operations. Furthermore, this critical review highlights the latest innovations in remediation techniques and various strategies to minimize airport waste. It shifts the paradigm of soil and water pollution management towards nature-based solutions, aligning with the sustainable development goals of the 2030 Agenda

  • Algebraic periods and minimal number of periodic points for smooth self-maps of 1-connected 4-manifolds with definite intersection forms
    • Haibao Duan
    • Grzegorz Graff
    • Jerzy Jezierski
    • Adrian Myszkowski
    2024 Full text Journal of Fixed Point Theory and Applications

    Let M be a closed 1-connected smooth 4-manifolds, and let r be a non-negative integer. We study the problem of finding minimal number of r-periodic points in the smooth homotopy class of a given map f: M-->M. This task is related to determining a topological invariant D^4_r[f], defined in Graff and Jezierski (Forum Math 21(3):491–509, 2009), expressed in terms of Lefschetz numbers of iterations and local fixed point indices of iterations. Previously, the invariant was computed for self-maps of some 3-manifolds. In this paper, we compute the invariants D^4_r[f] for the self-maps of closed 1-connected smooth 4-manifolds with definite intersection forms (i.e., connected sums of complex projective planes). We also present some efficient algorithmic approach to investigate that problem.

  • Alginate-based sorbents in miniaturized solid phase extraction techniques - Step towards greenness sample preparation
    • Natalia Jatkowska
    2024 Full text TRAC-TRENDS IN ANALYTICAL CHEMISTRY

    In response to growing concerns about environmental degradation, one of the main areas of research activity in recent years has been to make sample preparation methods more sustainable and eco-friendly. The increasing greenness of this step can be achieved by minimizing the usage of reagents, automating individual stages, saving energy and time, and using non-toxic, biodegradable substances. Therefore, the use of natural materials as sorbents in miniaturized extraction techniques is becoming a main trend. One of the natural material that is increasingly being used, not only due to eco-friendly nature but also because of their easy applicability to various sample preparation techniques, is alginate hydrogel. Following this trend, this review discusses the recent application of alginate-based sorbents in various microextraction techniques, focusing on functionalization approaches that enhance extraction performance. Additionally, the green profile of alginate-based sorbent microextraction approaches, along with the sorbent synthesis, were investigated.

  • Algorithmic Human Resources Management
    • Łukasz Sienkiewicz
    2024

    The rapid evolution of Digital Human Resources Management has introduced a transformative era where algorithms play a pivotal role in reshaping the landscape of workforce management. This transformation is encapsulated in the concepts of algorithmic management and algorithmic Human Resource Management (HRM). The integration of advanced analytics, predictive and prescriptive analytics and the power of Artificial Intelligence (AI) has given rise to algorithmic approaches that go beyond traditional human-centric decision-making. This paradigm shift is marked by comprehensive data collection, real-time responses to data influencing management decisions and automated decision-making processes. The culmination of algorithmic management is illustrated in a scheme where human-configured management processes are automated, leading to autonomous decision-making by an information system. The ongoing development of AI-driven algorithms, adapting and becoming self-learning, raises the prospect of increased automation, potentially leading to the displacement of human managers. This chapter provides a comprehensive overview of the evolution of algorithmic management and algorithmic HRM, setting the stage for a deeper exploration of their implications on organisational decision-making, employee management and the future of algorithm-supported management. While algorithmic HRM brings benefits to organisations it also poses risks, such as hidden errors and ethical concerns, emphasising the need for transparency and responsible application. Unconscious system errors, especially in fully automated systems, can lead to detrimental personnel decisions. Organisations must actively counteract potential negative consequences, striking a balance between technology and human decision-making, fostering organisational culture that embraces digital transformation while prioritising the human element. Enterprises should view technology as a tool to enhance work processes, focusing on connecting people and technology to create an innovative and inclusive work environment.

  • Alkyl polyglycoside-assisted separation followed by smartphone-based digital image colorimetry for on-site determination of total phenolic content in plant-based milk alternatives
    • Lutfi Yahya
    • Marek Tobiszewski
    • Khrystyna Vakh
    2024 MICROCHEMICAL JOURNAL

    A low-cost lab-on-a-smartphone platform for rapid and on-site assessment of total phenolic content in plant-based milk alternatives has been proposed for the first time. The three main components, including a smartphone, a disposable Eppendorf vial and a specially designed holder with an integrated polychromatic light source, were compactly assembled to enable color-forming reaction, separation and measurement. The method involves the application of the classically used Folin-Ciocalteu reagent to perform a color-forming reaction and the green surfactant alkyl polyglycoside C8-C10 to separate derivatives and impurities of the sample matrix. The coacervation phenomenon proceeding due to the addition of heptanoic acid to a micellar solution facilitates efficient phase separation and overcomes the challenges posed by matrix components such as ash, protein, carbohydrates and crude fibre. It also eliminates the need to use centrifugation and filtration as the matrix components were isolated into the obtained supramolecular solvent. The digital images were captured inside a lab-on-a-smartphone platform with controlled light conditions and analyzed using image processing algorithms. Various types of plant-based milk alternatives, including extracts from cereals, legumes, nuts, seeds and pseudocereals, were analyzed and the total phenolic content was found in the range 34.6 to 2748.7 mg GAE L−1. The results obtained, compared with the reference method, demonstrate significant findings, with a slope of the Passing-Bablok regression equal to 1.0049, indicating results are in good agreement. The linear range for total phenolic content was established as 25–110 mg GAE L−1, with a coefficient of determination of 0.9957. The limit of quantification and limit of detection were determined to be 25 mg GAE L−1 and 15 mg GAE L−1, respectively. Inter-day and intra-day precision, expressed as coefficient of variation (% CV), were 7.28 % and 2.29 %, respectively, while recovery rates ranged from 73 % to 154 %. Additionally, a greenness assessment using the AGREE tool showed an overall score of 0.84, indicating that the proposed smartphone-based method has a low environmental footprint. The total analysis time did not exceed 10 min, which is satisfactory for on-site analysis.

  • All but one expanding Lorenz maps with slope greater than or equal to $\sqrt 2$ are leo
    • Piotr Bartłomiejczyk
    • Piotr Nowak-Przygodzki
    2024 Colloquium Mathematicum

    We prove that with only one exception, all expanding Lorenz maps $f\colon[0,1]\to[0,1]$ with the derivative $f'(x)\ge\sqrt{2}$ (apart from a finite set of points) are locally eventually onto. Namely, for each such $f$ and each nonempty open interval $J\subset(0,1)$ there is $n\in\N$ such that $[0,1)\subset f^n(J)$. The mentioned exception is the map $f_0(x)=\sqrt{2}x+(2-\sqrt{2})/2 \pmod 1$. Recall that $f$ is an expanding Lorenz map if it is strictly increasing on $[0,c)$ and $[c,1]$ for some $c$ and satisfies the condition $\inf{f'}>1$.

  • Alphitobius diaperinus larvae (lesser mealworm) as human food – An approval of the European Commission – A critical review
    • Shahida Anusha Siddiqui
    • Y.s. Wu
    • K. Vijeepallam
    • K. Batumalaie
    • M.h.m. Hatta
    • H. Lutuf
    • Roberto Castro Munoz
    • I. Fernando
    • S.a. Ibrahim
    2024 Full text Journal of Insects as Food and Feed

    Due to the increasing threat of climate change and the need for sustainable food sources, human consumption of edible insects or entomophagy has gained considerable attention globally. The larvae of Alphitobius diaperinus Panzer (Coleoptera: Tenebrionidae), also known as the lesser mealworm, have been identified as a promising candidate for mass-rearing as a food source based the on evaluation on several aspects such as the production process, the microbiological and chemical composition, and the potential allergenicity to humans. As a consequence, the European Commission has recently approved the utilization of lesser mealworms as human foods. Lesser mealworms are considered a good source of protein, with a protein content ranging from 50-65% of their dry weight and containing various essential amino acids. Lesser mealworms are also rich in other essential nutrients such as iron, calcium, and vitamins B12 and B6. Furthermore, the hydrolysates of lesser mealworms are known to contain antioxidants, suggesting the therapeutic properties of the insects. To enable and ensure a continuous supply of lesser mealworms, various rearing procedures of the insects and information on optimal environmental rearing conditions have been reported. However, like other edible insects, lesser mealworms are still not commonly consumed in Western countries because of various consumer- and product-related factors. Ultimately, the European Commission’s approval of lesser mealworms as a novel food is a key milestone in the development of the insect food industry. Embracing the consumption of edible insects can help address the challenges of feeding a growing population, mitigate the environmental impact of food production, and promote a more sustainable and resilient food system for the future.

  • An absorbing set for the Chialvo map
    • Paweł Pilarczyk
    • Grzegorz Graff
    2024 Communications in Nonlinear Science and Numerical Simulation

    The classical Chialvo model, introduced in 1995, is one of the most important models that describe single neuron dynamics. In order to conduct effective numerical analysis of this model, it is necessary to obtain a rigorous estimate for the maximal bounded invariant set. We discuss this problem, and we correct and improve the results obtained by Courbage and Nekorkin (2010). In particular, we provide an explicit formula for an absorbing set for the Chialvo neuron model. We also introduce the notion of a weakly absorbing set, outline the methodology for its construction, and show its advantage over an absorbing set by means of numerical computations.

  • An Adaptive Network Model for a Double Bias Perspective on Learning from Mistakes within Organizations
    • Mojgan Hosseini
    • Jan Treur
    • Wioleta Kucharska
    2024

    Although making mistakes is a crucial part of learning, it is still often being avoided in companies as it is considered as a shameful incident. This goes hand in hand with a mindset of a boss who dominantly believes that mistakes usually have negative consequences and therefore avoids them by only accepting simple tasks. Thus, there is no mechanism to learn from mistakes. Employees working for and being influenced by such a boss also strongly believe that mistakes usually have negative consequences but in addition they believe that the boss never makes mistakes, it is often believed that only those who never make mistakes can be bosses and hold power. That’s the problem, such kinds of bosses do not learn. So, on the one hand, we have bosses who select simple tasks to be always seen as perfect. Therefore, also they believe they should avoid mistakes. On the other hand, there exists a mindset of a boss who is not limited to simple tasks, he/she accepts more complex tasks and therefore in the end has better general performance by learning from mistakes. This then also affects the mindset and actions of employees in the same direction. This paper investigates the consequences of both attitudes for the organizations. It does so by computational analysis based on an adaptive dynamical systems modeling approach represented in a network format using the self-modeling network modeling principle.

  • An Adversarial Machine Learning Approach on Securing Large Language Model with Vigil, an Open-Source Initiative
    • Kushal Pokhrel
    • Cesar Sanín
    • Md Rafiqul Islam
    • Edward Szczerbicki
    2024 Full text Procedia Computer Science

    Several security concerns and efforts to breach system security and prompt safety concerns have been brought to light as a result of the expanding use of LLMs. These vulnerabilities are evident and LLM models have been showing many signs of hallucination, repetitive content generation, and biases, which makes them vulnerable to malicious prompts that raise substantial concerns in regard to the dependability and efficiency of such models. It is vital to have a complete grasp of the complex behaviours of malicious attackers in order to build effective strategies for protecting modern artificial intelligence (AI) systems through the development of effective tactics. The purpose of this study is to look into some of these aspects and propose a method for preventing devastating possibilities and protecting LLMs from potential threats that attackers may pose. Vigil is an open-source LLM prompt security scanner, that is accessible as a Python library and REST API, specifically to solve these problems by employing a sophisticated adversarial machine-learning algorithm. The entire objective of this study is to make use of Vigil as a security scanner. and asses its efficiency. In this case study, we shed some light on Vigil, which effectively recognises and helps LLM prompts by identifying two varieties of threats: malicious and benign.

  • An Analysis of Airline GRI and SDG Reporting
    • Eljas Johansson
    2024 Full text

    This study aims to increase our understanding of the Global Reporting Initiative’s (GRI) topic-specific disclosures and the sustainable development goals (SDGs) addressed in the global passenger airline industry’s sustainability reporting (SR). Based on a quantitative content analysis of the industry’s sustainability reports from the financial year 2019 (FY19), this study reveals that airlines focused more on reporting environmental issues, especially emissions, than economic or social dimensions, demonstrating this emission-intensive industry’s responsiveness to stakeholders’ information needs. However, a closer look at the reported impacts shows that many topic-specific disclosures and SDGs, which industry associations have not identified as relevant to the industry, were also mentioned across the reports. Moreover, the results indicated a broader use of SDGs in Asia-Pacific reports than in European. The results are expected to interest practitioners and academics in assessing and developing the industry’s SR.

  • An Analysis of the Performance of Lightweight CNNs in the Context of Object Detection on Mobile Phones
    • Jakub Łęcki
    • Marek Hering
    • Maciej Jabłoński
    • Aleksandra Karpus
    2024

    Convolutional Neural Networks (CNNs) are widely used in computer vision, which is now increasingly used in mobile phones. The problem is that smartphones do not have much processing power. Initially, CNNs focused solely on increasing accuracy. High-end computing devices are most often used in this type of research. The most popular application of lightweight CNN object detection is real-time image processing, which can be found in devices such as cameras and autonomous vehicles. Therefore, there is a need to optimize CNNs for use on mobile devices. This paper presents the comparision of latency and mAP of 22 lightweight CNN models from the MobileNet and EfficientDet families measured on 7 mobile phones.