Hasil untuk "Industry"

Marta A. Geletkanycz, D. Hambrick

Duncan Leeson, N. M. Dowell, N. Shah et al.

Abstract In order to meet the IPCC recommendation for an 80% cut in CO2 emissions by 2050, industries will be required to drastically reduce their emissions. To meet these targets, technologies such as carbon capture and storage (CCS) must be part of the economic set of decarbonisation options for industry. A systematic review of the literature has been carried out on four of the largest industrial sectors (the iron and steel industry, the cement industry, the petroleum refining industry and the pulp and paper industry) as well as selected high-purity sources of CO2 from other industries to assess the applicability of different CCS technologies. Costing data have been gathered, and for the cement, iron and steel and refining industries, these data are used in a model to project costs per tonne of CO2 avoided over the time period extending from first deployment until 2050. A sensitivity analysis was carried out on the model to assess which variables had the greatest impact on the overall cost of wide-scale CCS deployment for future better targeting of cost reduction measures. The factors found to have the greatest overall impact were the initial cost of CCS at the start of deployment and the start date at which large scale deployment is started, whilst a slower initial deployment rate after the start date also leads to significantly increased costs.

Jie Xiong, Degui Wang, Liping Xie et al.

The construction of mass concrete foundations for nuclear power plants faces significant challenges in controlling hydration heat and preventing early-age thermal cracking. This study develops an integrated framework combining high-fidelity thermal–mechanical simulation, real-time temperature monitoring, and construction process optimization to address these issues. Focusing on the VVER-1200 reactor raft foundation in the Xudapu NPP Phase II Project, an innovative center-to-periphery synchronous pouring method is proposed, departing from conventional inclined or layered pouring by strategically utilizing stage time lags to moderate the radial temperature gradient. Numerical simulations demonstrate that this method significantly reduces the peak temperature and thermal stress. Field validation shows that the maximum core-to-surface temperature difference is controlled within 19.8 °C, well below the critical threshold of 25 °C, and the peak concrete temperature remains at 66.7 °C, safely below the risk level for delayed ettringite formation (82–85 °C). The cracking risk coefficient K remains below 0.65, indicating a low probability of thermal cracking. Post-construction inspection confirms the absence of thermal cracks in the 5240 m³ monolithic pour. The proposed methodology offers a reliable, science-based approach for thermal crack mitigation and serves as a valuable reference for similar large-scale mass concrete structures in nuclear and other critical infrastructure projects.

Najla A. Barnawi, Fay A. AlAmmar, Sultan A. Aldabeis et al.

Abstract Despite the growing role of AI and robotics in healthcare, little is known about their integration into dental care for persons with disabilities (PWDs) in Saudi Arabia. This study aimed to assess dentists’ perceptions and attitudes towards and use of RT/AI in dentistry and identify the predictors of using RT/AI to care for PWDs in the Saudi context. A cross-sectional study was conducted using a previously validated online self-reported questionnaire via SurveyMonkey, targeting 309 Saudi and non-Saudi licensed dentists and dental/oral health practitioners, to collect data on the following: 1) Personal and work-related characteristics, 2) Perception toward RT/AI use, 3) Attitude toward using AI and RT in dentistry, and 4) Current use of RT and AI. RT/AI use rate was calculated for each clinical aspect and each type of impairment. Logistic regression analysis was used to identify the predictors of dentists’ use of RT and AI to provide care for PWDs. Significance was set at p < 0.05. Our study revealed that 59.2% of dentists who worked with PWDs reported utilizing RT/AI in various clinical aspects. Almost one-fourth of dentists reported using RT/AI in clinical examinations (23.9%), managing complications (26.8%), and performing invasive procedures (28.6%). Nearly one-third of respondents reported using RT/AI for taking a history (30%), non-invasive procedures (31.5%), behavioral training sessions (32.9%), health education (36.2%), medical diagnosis (36.6%), diagnostic tests (38%), and treatment planning (43.7%). Over one-half (54.9%) and one-fourth (28.6%) of the dentists reported a positive perception and attitude towards RT/AI use in dentistry. However, after adjusting for possible confounders, only previous RT/AI training remained a significant predictor of RT/AI use among dentists working with PWDs (OR = 9.18, 95% CI 2.92–28.90, p < 0.001). Our study is the first in the Saudi context to investigate the use of RT and AI by dentists caring for PWDs. Previous training was associated with greater use of RT/AI in this context. Potential collaborations between dental institutes and stakeholders in the RT and AI industry are recommended.

Yang HE, Yiting FENG, Haowen WU et al.

To improve traditional sweet rice wine stability and flavor, yeasts with high alcohol and ester production were isolated, purified, and screened from Yunnan traditional sweet rice wine, and the strain was identified by morphological observation and molecular biology methods, with its alcohol and ester production characteristics analyzed. Effects of different yeasts synergized with Rhizopus oryzae on the quality of sweet rice wine was assessed by physicochemical indexes, sensory evaluation, and amino acid and volatile compositions. Results showed that a high alcohol production yeast strain LF05 and a high ester production yeast strain PL10 were isolated and screened, which were identified as Saccharomyces cerevisiae and Rhodotorula mucilaginosa. Compared with the control, S. cerevisiae LF05 synergised with Rhizopus oryzae fermented sweet rice wine increased the alcoholic strength (6.80vol%) and volatile alcohols (44.27 mg/L) by 23.64% and 36.60%. It reduced the bitter amino acids (329.25 μg/mL) by 41.19%. Compared with the control group Rhodotorula mucilaginosa PL10 synergized synergised with Rhizopus oryzae fermented sweet rice wine volatile esters (7.29 mg/L), sweet amino acids (425.18 μg/mL), and fresh amino acids (740.51 μg/mL) were increased by 220.39%, 154.64%, and 134.82%, as well as the sensory score (86 points) increased by 64.48%. In conclusion, the synergistic fermentation of S. cerevisiae LF05 and R. mucilaginosa PL10 with Rhizopus oryzae can improve the problem of taste and blandness of commercial sweet rice wine, which provides a reference and theoretical basis for the preparation of sweet rice wine in Yunnan Province.

Myung-Su Yi, Joo-Shin Park

The living quarters (LQ) on jack-up rigs play a critical role in ensuring crew safety and operational functionality under extreme offshore conditions. This study presents a comprehensive structural engineering procedure for the design and analysis of LQ structures, addressing the absence of specific industry standards. The methodology integrates global and local load effects from critical equipment, such as helidecks and lifeboat stations, under harsh environmental conditions during wet towing. A multi-level analysis approach, including finite element analysis (FEA), nonlinear evaluations, and fatigue assessments, was employed to verify structural resilience. The study successfully validates the LQ structures against ultimate limit state (ULS), serviceability limit state (SLS), and accidental limit state (ALS) criteria. The maximum plastic strain observed under green water pressure was 3.8 %, well below the allowable threshold of 15 %, demonstrating adequate safety margins. Fatigue analysis confirmed resistance to vortex-induced vibrations (VIV), ensuring the durability of tubular members. Optimization efforts reduced LQ structural weight by 20 %, enhancing efficiency without compromising safety. The proposed procedure bridges the gap in industry standards, providing a robust framework for designing safer and more reliable LQ structures. This study advances offshore engineering practices by addressing complex loading scenarios and operational challenges, thereby supporting the development of resilient jack-up rigs capable of enduring extreme marine conditions.

Liza Efriyanti, Ihwana As'ad

The design of curricula in Islamic universities frequently encounters difficulties in addressing the evolving needs of students, industry demands and the distinctive integration of Islamic values. Conventional methodologies are inadequate in their capacity to adapt to the evolving needs of the modern educational landscape. Furthermore, the integration of artificial intelligence (AI) in this domain remains underdeveloped, with many instances overlooking the crucial role of religious principles and institutional characteristics. This study addresses this gap by developing a Decision Support System (DSS) using Mamdani type 1 fuzzy logic, with the objective of assisting in determining an independent curriculum learning model tailored to Islamic higher education. The system incorporates a number of input variables, including student needs, industry requirements, institutional characteristics and data analysis. The output variables include an evaluation of the suitability of the learning model and a recommendation as to the most appropriate model. To illustrate, in situations where student needs are high, industry demands are moderate, institutional characteristics are high, and data analysis is moderate, the recommended model places an emphasis on balancing theoretical knowledge with practical application, while also aligning with Islamic values.  The validation of this AI-based model, utilizing 2023 historical data from five Islamic universities in West Sumatra, yielded an average Mean Absolute Error (MAE) of 0.64, thereby demonstrating good predictive accuracy. The integration of AI in this system facilitates data-driven decision-making, thereby enhancing the relevance and adaptability of the curriculum. It has the potential to improve the quality of education, support balanced student learning outcomes, and ensure alignment with Islamic principles, making it a transformative tool for curriculum development in Islamic higher education.

Ahmad M.Zamil, Mohammad Alhusban, Alharkan Abdulrahman

Abstract This research rigorously examines the implementation challenges of value management (VM) implementation challenges within construction projects across Jordan. This study is guided by a conceptual framework linking organisational factors to VM implementation barriers, highlighting the impact on project outcomes and providing a basis for targeted interventions in developing sustainable construction. Utilizing a comprehensive survey, insights were gathered from 103 industry experts, and the collected data were rigorously analyzed through exploratory factor analysis (EFA) and partial least squares structural modeling (PLS-SEM). The analysis identified four critical constructs of VM implementation challenges: culture, awareness, resources, and policy. The analysis identified four important constructs of VM adoption challenges: culture, awareness, resources, and policy. Among these constructs, culture appeared to be the most important barrier (path-coefficient = 0.317), followed by resources (0.285), awareness (0.271), and policy (0.209). The findings not only illuminate the complex dynamics within Jordan's construction sector but also offer valuable implications for other developing nations with comparable socio-economic and environmental backgrounds, struggling with similar challenges. This study contributes substantially to enhancing the understanding among stakeholders of the barriers to effective VM implementation and proposes actionable strategies to mitigate these challenges, ultimately aiming to improve construction quality and cost-efficiency in developing contexts. However, this research is limited to Jordan; thus, future studies covering a broader geographical scope are recommended.

Rafael da Silva Maciel, Lucio Veraldo

The evolution of the 5S methodology with the support of artificial intelligence techniques represents a significant opportunity to improve industrial organization audits in the automotive chain, making them more objective, efficient and aligned with Industry 4.0 standards. This work developed an automated 5S audit system based on large-scale language models (LLM), capable of assessing the five senses (Seiri, Seiton, Seiso, Seiketsu, Shitsuke) in a standardized way through intelligent image analysis. The system's reliability was validated using Cohen's concordance coefficient (kappa = 0.75), showing strong alignment between the automated assessments and the corresponding human audits. The results indicate that the proposed solution contributes significantly to continuous improvement in automotive manufacturing environments, speeding up the audit process by 50% of the traditional time and maintaining the consistency of the assessments, with a 99.8% reduction in operating costs compared to traditional manual audits. The methodology presented establishes a new paradigm for integrating lean systems with emerging AI technologies, offering scalability for implementation in automotive plants of different sizes.

Katharina Ledebur. Ladislav Bartuska, Klaus Friesenbichler, Peter Klimek

The automotive industry is undergoing transformation, driven by the electrification of powertrains, the rise of software-defined vehicles, and the adoption of circular economy concepts. These trends blur the boundaries between the automotive sector and other industries. Unlike internal combustion engine (ICE) production, where mechanical capabilities dominated, competitiveness in electric vehicle (EV) production increasingly depends on expertise in electronics, batteries, and software. This study investigates whether and how firms' ability to leverage cross-industry diversification contributes to competitive advantage. We develop a country-level product space covering all industries and an industry-specific product space covering over 900 automotive components. This allows us to identify clusters of parts that are exported together, revealing shared manufacturing capabilities. Closeness centrality in the country-level product space, rather than simple proximity, is a strong predictor of where new comparative advantages are likely to emerge. We examine this relationship across industrial sectors to establish patterns of path dependency, diversification and capability formation, and then focus on the EV transition. New strengths in vehicles and aluminium products in the EU are expected to generate 5 and 4.6 times more EV-specific strengths, respectively, than other EV-relevant sectors over the next decade, compared to only 1.6 and 4.5 new strengths in already diversified China. Countries such as South Korea, China, the US and Canada show strong potential for diversification into EV-related products, while established producers in the EU are likely to come under pressure. These findings suggest that the success of the automotive transformation depends on regions' ability to mobilize existing industrial capabilities, particularly in sectors such as machinery and electronic equipment.

Joseph Heeley, Samuel Boobier, Jonathan D. Hirst

Abstract Selecting greener solvents during experiment design is imperative for greener chemistry. While many solvent selection guides are currently used in the pharmaceutical industry, these are often paper-based guides which can make it difficult to identify and compare specific solvents. This work presents a stand-alone version of the solvent flashcards that were developed as part of the AI4Green electronic laboratory notebook. The functionality is an intuitive and interactive interface for the visualisation of data from CHEM21, a pharmaceutical solvent selection guide that categorises solvents according to “greenness”. This open-source software is written in Python, JavaScript, HTML and CSS and allows users to directly contrast and compare specific solvents by generating colour-coded flashcards. It can be installed locally using pip, or alternatively the source code is available on GitHub: https://github.com/AI4Green/solvent_flashcards . The documentation can also be found on GitHub or on the corresponding Python Package Index webpage: https://pypi.org/project/solvent-guide/ . Scientific Contribution This simple and easy-to-use digital tool provides a visualisation of solvent greenness data through a novel intuitive interface and encourages green chemistry. It offers numerous advantages over traditional solvent selection guides, allowing users to directly customise the solvent list and generate side-by-side comparisons of only the most important solvents. The release as a standalone package will maximise the benefit of this software. Graphical Abstract

Kurukulasooriya Fernando ana Gianluca Demartini

Recent advancements of generative LLMs (Large Language Models) have exhibited human-like language capabilities but have shown a lack of domain-specific understanding. Therefore, the research community has started the development of domain-specific LLMs for many domains. In this work we focus on discussing how to build mining domain-specific LLMs, as the global mining industry contributes significantly to the worldwide economy. We report on MiningGPT, a mining domain-specific instruction-following 7B parameter LLM model which showed a 14\% higher mining domain knowledge test score as compared to its parent model Mistral 7B instruct.

Juliette Grosset, Alain-Jérôme Fougères, M Djoko-Kouam et al.

One of the challenges of Industry 4.0, is to determine and optimize the flow of data, products and materials in manufacturing companies. To realize these challenges, many solutions have been defined such as the utilization of automated guided vehicles (AGVs). However, being guided is a handicap for these vehicles to fully meet the requirements of Industry 4.0 in terms of adaptability and flexibility: the autonomy of vehicles cannot be reduced to predetermined trajectories. Therefore, it is necessary to develop their autonomy. This will be possible by designing new generations of industrial autonomous vehicles (IAVs), in the form of intelligent and cooperative autonomous mobile robots.In the field of road transport, research is very active to make the car autonomous. Many algorithms, solving problematic traffic situations similar to those that can occur in an industrial environment, can be transposed in the industrial field and therefore for IAVs. The technologies standardized in dedicated bodies (e.g., ETSI TC ITS), such as those concerning the exchange of messages between vehicles to increase their awareness or their ability to cooperate, can also be transposed to the industrial context. The deployment of intelligent autonomous vehicle fleets raises several challenges: acceptability by employees, vehicle location, traffic fluidity, vehicle perception of changing environments (dynamic), vehicle-infrastructure cooperation, or vehicles heterogeneity. In this context, developing the autonomy of IAVs requires a relevant working method. The identification of reusable or adaptable algorithms to the various problems raised by the increase in the autonomy of IAVs is not sufficient, it is also necessary to be able to model, to simulate, to test and to experiment with the proposed solutions. Simulation is essential since it allows both to adapt and to validate the algorithms, but also to design and to prepare the experiments.To improve the autonomy of a fleet, we consider the approach relying on a collective intelligence to make the behaviours of vehicles adaptive. In this chapter, we will focus on a class of problems faced by IAVs related to collision and obstacle avoidance. Among these problems, we are particularly interested when two vehicles need to cross an intersection at the same time, known as a deadlock situation. But also, when obstacles are present in the aisles and need to be avoided by the vehicles safely.

Peter Klimek, Maximilian Hess, Markus Gerschberger et al.

The steel industry is a major contributor to CO2 emissions, accounting for 7% of global emissions. The European steel industry is seeking to reduce its emissions by increasing the use of electric arc furnaces (EAFs), which can produce steel from scrap, marking a major shift towards a circular steel economy. Here, we show by combining trade with business intelligence data that this shift requires a deep restructuring of the global and European scrap trade, as well as a substantial scaling of the underlying business ecosystem. We find that the scrap imports of European countries with major EAF installations have steadily decreased since 2007 while globally scrap trade started to increase recently. Our statistical modelling shows that every 1,000 tonnes of EAF capacity installed is associated with an increase in annual imports of 550 tonnes and a decrease in annual exports of 1,000 tonnes of scrap, suggesting increased competition for scrap metal as countries ramp up their EAF capacity. Furthermore, each scrap company enables an increase of around 79,000 tonnes of EAF-based steel production per year in the EU. Taking these relations as causal and extrapolating to the currently planned EAF capacity, we find that an additional 730 (SD 140) companies might be required, employing about 35,000 people (IQR 29,000-50,000) and generating an additional estimated turnover of USD 35 billion (IQR 27-48). Our results thus suggest that scrap metal is likely to become a strategic resource. They highlight the need for a massive restructuring of the industry's supply networks and identify the resulting growth opportunities for companies.

Atif Hussain, Rana Rizwan

This paper argues for the strategic treatment of artificial intelligence as a key industry within broader industrial policy framework of Pakistan, underscoring the importance of aligning it with national goals such as economic resilience and preservation of autonomy. The paper starts with defining industrial policy as a set of targeted government interventions to shape specific sectors for strategic outcomes and argues for its application to AI in Pakistan due to its huge potential, the risks of unregulated adoption, and prevailing market inefficiencies. The paper conceptualizes AI as a layered ecosystem, comprising foundational infrastructure, core computing, development platforms, and service and product layers, supported by education, government policy, and research and development. The analysis highlights that AI sector of Pakistan is predominantly service oriented, with limited product innovation and dependence on foreign technologies, posing risks to economic independence, national security, and employment. To address these challenges, the paper recommends educational reforms, support for local AI product development, initiatives for indigenous cloud and hardware capabilities, and public-private collaborations on foundational models. Additionally, it advocates for public procurement policies and infrastructure incentives to foster local solutions and reduce reliance on foreign providers. This strategy aims to position Pakistan as a competitive, autonomous player in the global AI ecosystem.

Jeong Kuk Kim, Byongug Jeong, Jae-Hyuk Choi et al.

This study aimed to evaluate the environmental impact of using liquefied petroleum gas (LPG) in small fishing vessels by conducting a life cycle assessment (LCA) in Korea. For the first time in the country, LPG engines designed for small fishing ships were utilized in this study. In addition, this research examined the potential benefits of employing Bio LPG, a renewable LPG produced from two distinct raw materials (crude palm oil (CPO) and refined, bleached, and deodorized (RBD) palm oil), instead of conventional LPG. The LCA findings reveal that utilizing LPG fuel in small fishing vessels can reduce greenhouse gas (GHG) emissions by more than 30% over conventional gasoline and diesel fuels. During the life cycle of vessels that use LPG fuel instead of gasoline and diesel fuels, there is a reduction of 2.2 and 1.2 million tons of GHG emissions, respectively. Moreover, substituting conventional fossil fuels with Bio LPG can result in over 65% reduction in GHG emissions. For the life cycle of boats that use Bio LPG fuel in place of gasoline and diesel fuels, the reduction of GHG emissions was 4.9 million tons and 2.5 million tons for CPO and 5.2 million tons and 2.7 million tons for RBD, respectively. This study not only underscores the substantial advantages of using Bio LPG over conventional fossil fuels but also presents conventional LPG as a way to reduce GHG emissions and promote sustainable practices in the fishing industry.

Saara Tenhunen, Tomi Männistö, Petri Ihantola et al.

Previous research has demonstrated that preparing students for life in software engineering is not a trivial task. Authentic learning experiences are challenging to provide, and there are gaps between what students have done at the university and what they are expected to master when getting into the industry after graduation. To address this challenge, we present a novel way of teaching industry-relevant skills in a university-led internal software startup called Software Development Academy (SDA). In addition to describing the SDA concept in detail, we have investigated what educational aspects characterise SDA and how it compares to capstone projects. The questions are answered based on 15 semi-structured interviews with alumni of SDA. Working with production-quality software and having a wide range of responsibilities were perceived as the most integral aspects of SDA and provided students with a comprehensive skill set for the future.

Anne de Bortoli, Maxime Agez

Industries struggle to build robust environmental transition plans as they lack the tools to quantify their ecological responsibility over their value chain. Companies mostly turn to sole greenhouse gas (GHG) emissions reporting or time-intensive Life Cycle Assessment (LCA), while Environmentally-Extended Input-Output (EEIO) analysis is more efficient on a wider scale. We illustrate EEIO analysis usefulness to sketch transition plans on the example of Canada s road industry - estimation of national environmental contributions, most important environmental issues, main potential transition levers of the sector, and metrics prioritization for green purchase plans). To do so, openIO-Canada, a new Canadian EEIO database, coupled with IMPACT World plus v1.30-1.48 characterization method, provides a multicriteria environmental diagnosis of Canada s economy. The road industry generates a limited impact (0.5-1.8 percent) but must reduce the environmental burden from material purchases - mainly concrete and asphalt products - through green purchase plans and eco-design and invest in new machinery powered with cleaner energies such as low-carbon electricity or bioenergies. EEIO analysis also captures impacts often neglected in process-based pavement LCAs - amortization of capital goods, staff consumptions, and services - and shows some substantial impacts advocating for enlarging system boundaries in standard LCA. Yet, pavement construction and maintenance only explain 5 percent of the life cycle carbon footprint of Canada s road network, against 95 percent for the roads usage. Thereby, a carbon-neutral pathway for the road industry must first focus on reducing vehicle consumption and wear through better design and maintenance of roads (...)

Lijuan Gao, Qi Yan, Jie Li et al.

To investigate the effects of the dietary inclusion of elephant grass on the growth performance, blood profiles, carcass characteristics, ileum and stomach microbiota of fattening pigs, pigs were fed one of seven diets including a basal diet (Control), and six treatments, where the basal diet was supplemented with 10%, 15% or 20% of elephant grass, Cenchrus purpureus cv. Guiminyin (CpGui10, CpGui15, CpGui20) or cv. Purple (CpP10, CpP15, CpP20). Results showed that supplementation of 20% CpGui in the diet significantly increased (P &lt; 0.05) average daily gain (ADG) and gain to feed (G/F) ratio by the end of the experiment. Additionally, pigs fed the CpGui20 diet showed higher (0.01 &lt; P &lt; 0.05) slaughter weight and tended to have increased loin-eye area and lean meat percentage, and, decreased backfat thickness compared with control pigs. Furthermore, 16S ribosomal DNA gene amplicon profiling showed that the inclusion of elephant grass in the diet was associated with modulation of the ileum and stomach microbiota composition at the order level. Relative abundance of the Lactobacillales order in the ileum and stomach increased with different proportions of elephant grass, while that of Enterobacteriales decreased. In conclusion, these results indicate that at up to 20% inclusion in the diet of pigs, elephant grass can promote enhanced growth performance and carcass characteristics, and, modulate the ileum and stomach microbiota composition of the pigs.

Chuyi Li, Lulu Li, Hongliang Jiang et al.

For years, the YOLO series has been the de facto industry-level standard for efficient object detection. The YOLO community has prospered overwhelmingly to enrich its use in a multitude of hardware platforms and abundant scenarios. In this technical report, we strive to push its limits to the next level, stepping forward with an unwavering mindset for industry application. Considering the diverse requirements for speed and accuracy in the real environment, we extensively examine the up-to-date object detection advancements either from industry or academia. Specifically, we heavily assimilate ideas from recent network design, training strategies, testing techniques, quantization, and optimization methods. On top of this, we integrate our thoughts and practice to build a suite of deployment-ready networks at various scales to accommodate diversified use cases. With the generous permission of YOLO authors, we name it YOLOv6. We also express our warm welcome to users and contributors for further enhancement. For a glimpse of performance, our YOLOv6-N hits 35.9% AP on the COCO dataset at a throughput of 1234 FPS on an NVIDIA Tesla T4 GPU. YOLOv6-S strikes 43.5% AP at 495 FPS, outperforming other mainstream detectors at the same scale~(YOLOv5-S, YOLOX-S, and PPYOLOE-S). Our quantized version of YOLOv6-S even brings a new state-of-the-art 43.3% AP at 869 FPS. Furthermore, YOLOv6-M/L also achieves better accuracy performance (i.e., 49.5%/52.3%) than other detectors with a similar inference speed. We carefully conducted experiments to validate the effectiveness of each component. Our code is made available at https://github.com/meituan/YOLOv6.