Ugo Ibusuki
@ufabc.edu.br
professor of management engineering
UFABC
RESEARCH, TEACHING, or OTHER INTERESTS
Industrial and Manufacturing Engineering, Automotive Engineering, Mechanical Engineering, Control and Systems Engineering
Scopus Publications
Scopus Publications
Ugo Ibusuki and Vinicius Mafra Viti
SAE International
<div class="section abstract"><div class="htmlview paragraph">The aerospace industry is undergoing a revolution with the large-scale development of eVTOL (Electric Vertical Take-Off &amp; Landing) and MEA (More Electric Aircraft). These aerial vehicles, many of them unmanned vehicles (UAV), will serve a variety of service-related functions: Search and Rescue (SAR), Medivac, delivery and lift operations, aerial mapping, and, of course, human transportation [<span class="xref">1</span>].</div><div class="htmlview paragraph">Despite its numerous functionalities, this type of vehicle has a serious problem, which is its usual batteries, the main means for its operation. Due to its autonomy not being so effective compared to its charging time, generating a considerable loss of time. In this context, it is necessary to find forms of components that can replace these batteries, so that the effective development of these vehicles is possible.</div><div class="htmlview paragraph">Studies done in other means of transportation point out that the use of hydrogen fuel cells has grown a lot. In this way, it is known that this type of fuel is seen as something of the future, but many companies have evaluated the possibility of implementation in cars, trains, and even airplanes. Therefore, a literature review will be presented about this new type of technology, demonstrating its advantages and its use. Finally, after careful analysis of the issues highlighted throughout the article, the best effective solution for replacing the batteries will be presented, a component that has caused major problems for the developers of this new aircraft [<span class="xref">28</span>].</div></div>
Renato Avila, Edson Kitani, Francisco De Assis Zampirolli, Leopoldo Yoshioka, Luiz Antonio Celiberto, and Ugo Ibusuki
IEEE
With the growing demand for efficiency in precision agriculture, the popularization of autonomous agricultural vehicles (AAV) is a trend. A safer operation is then necessary, due to the risk that the operation of these machines presents. Several sensors are used for this purpose, working together with computer vision and machine learning algorithms. The literature produced on the subject is vast and rapidly updated. However, different neural network architectures are available and can be tested in embedded systems for agricultural au-tomation. This article presents comparisons between different versions of YOLO and SSD Mobilenet networks, using 82 minutes of videos recorded in the context of a sugarcane harvest. These comparisons were performed on an AGX Jetson Orin and a desktop computer with an RTX 3060 Ti graphics card. In the first case, we also compare networks that use the library provided by the manufacturer (Jetson Inference) and also networks that use the OpenCV and PyTorch libraries. For an embedded system using AGX, the best results were found with the YoloV4 network provided by the OpenCV library. Better results may be found when more advanced networking architectures are available in native AGX Jetson Orin libraries.
Renato Avila, Edson Kitani, Francisco De Assis Zampirolli, Leopoldo Yoshioka, Luiz Antonio Celiberto, and Ugo Ibusuki
IEEE
Com o aumento da demanda por eficiência na agricultura de precisão, a popularização do uso de veículos agrícolas autônomos (VAAs) é uma tendência crescente. A literatura sobre o assunto é vasta e se atualiza rapidamente. O objetivo deste trabalho é analisar o conhecimento produzido sobre VAAs para identificar as melhores técnicas e estratégias de visão computacional e redes neurais artificiais. Foi realizada uma revisão sistemática abrangendo os últimos 10 anos para identificar publicações científicas relevantes cuja implementação foi realizada em veículos agrícolas autônomos. Foram encontrados excelentes resultados em 35 artigos analisados. Dentre eles, as soluções que se utilizam de redes neurais convolucionais se destacam, sendo as mais populares nos últimos anos, e com abordagens variadas, como aprendizado por transferência e redes que fazem uso de concatenações de sensores.
Ugo Ibusuki, Ailton Conde Jussani, Renan Degaspari d, e Araújo, Rafaela d, and e Paula Koto
Inderscience Publishers
Bruno B. Furquim, Ugo Ibusuki, Uriel G. Janota, and Gerhard Ett
SAE International
Nelson da Silva Bento, William Cavalcanti Bortoleto, and Ugo Ibusuki
SAE International
Bruno Bueno Furquim, Gustavo Morales Vieira, Pedro Pinto Araujo, Vinicius Mafra Viti, Rafaela Teixeira de Queiroz Agostinho, and Ugo Ibusuki
SAE International
Ugo Ibusuki, Paulo Carlos Kaminski, and Roberto Carlos Bernardes
Informa UK Limited
ABSTRACT The Brazilian Automotive and Aeronautic Industries are both important sectors for the National Innovation System. Through this research, we aim to understand the Innovation System of both sectors in order to collect lessons learned for the Automotive Industry Innovation Policy called Inovar-Auto (Innovate-Auto), which was effective from the beginning of 2013 until the end of 2017. This study was conducted based on the quantitative analysis of sectoral investment in R&D and the number of patents as well as on the qualitative analysis of some critical projects performed by companies in both industry sectors, the Automotive and the Aeronautic. All data were collected via interviews with the agents of the sectoral innovation system. This article aims to contribute to the analysis of a sectoral innovation system based on an analytical tool to measure the maturity level on specific key variables.
Erik Telles Pascoal, Maurício César Delamaro, Ugo Ibusuki, Osamu Tsukada, and Henrique Martins Rocha
Inderscience Publishers
To enhance the competitiveness of the Brazilian automotive industry, the government approved a new automotive policy in October 2012. It is called the Inovar-Auto program and will be implemented over the course of five years from 2013 to 2017. This program encourages automakers and their suppliers to produce more efficient, technologically advanced vehicles and auto parts by investing in R&D and engineering. The objective of this paper is to determine whether the Inovar-Auto program will effectively improve the competitiveness of the Brazilian automotive industry. In this sense, the impact of each Inovar-Auto requirement in relation to the Porter's diamond model and flowchart approach model was investigated. Although the Inovar-Auto program represents an advance over previous industrial policies, particularly in relation to aspects concerning improvement in energy efficiency, encouraging innovation, increased local content, the traceability of auto-parts and supplier development policy, the result of this research indicates that its implementation has not been stimulating a substantial evolution on the competitiveness of the Brazilian automotive industry.
Ailton Conde Jussani, Andreas Heer, Ugo Ibusuki, and Carlos de Moura Côrtes
SAE International
Ugo Ibusuki, Roberto C. Bernardes, and Flávia L. Consoni
Inderscience Publishers
This study analyses the Brazilian auto-industry based on Porter's diamond model, looking for the four competitive advantage factors and how the new automotive industrial policy (INOVAR-AUTO), recently launched in 2012, impacts on the local automobile industry. Results show that the Brazilian government's goal is in line with Porter's theoretical model, but it leads us to conclude that Brazil needs to devise, design and implement a focused and integrated industrial policy of research and development (R%D) in the country. The survey, conducted with Asian new comer car and auto-part manufacturers, shows they have quickly adapted to the new policy changes of INOVAR-AUTO without consolidating a real local R%D function. The absence of such a technological specialisation policy increases the chances for Brazil to become a centre of production and commercialisation of vehicles conceived externally and produced with increasingly less local content and local design with higher prices due to less competition.
Ugo Ibusuki, Hideo Kobayashi, and Paulo Carlos Kaminski
Inderscience Publishers
As global competition intensifies, transnational corporations are internationalising even the most knowledge-intensive corporate functions, such as research and development (R%D). Today, corporations in industries such as automobiles are establishing R%D facilities in selected developing countries. They do this to enhance their efficiency, to access expanding pools of scientists and engineers, and to meet the demands in these countries. These recent trends have important implications for the international division of labour and stress the need for coherent national policies to ensure benefits from this evolution. This study aims to identify the localisation of product development activities based on Porter's (1998) diamond model for competitive advantage of location, and Kuchiki's (2005) flowchart approach for government policies. The analysis is based on case studies of automobile assemblers that have headquarters in Triad markets and facilities in Brazil. Brazilian competitive advantages of location are investigated, and implications for host government policies proposed, from the viewpoint of leveraging location specific advantages for R%D localisation.
Ugo Ibusuki and Paulo Carlos Kaminski
Elsevier BV