Increased instrument intelligence--can it reduce laboratory error?

Thermocompression bonding is a unique process in semiconductor manufacturing since it essentially bridges the manufacturing space between front end semiconductor processes and back end of line packaging. The amount of potential data generated from a TCB process is enormous and the opportunities for process monitoring and control are significant. In a conventional mass reflow process temperature and oxygen levels can be tracked and periodic monitors can be run through the oven to verify profiles, but process data for individual die is not available. In a local reflow TCB process there is full material tracking back to the source wafer with die alignment, temperature, force, time to melt, stack layer position, and placement position logged for each die processed. Excursions in the process can be detected quickly and with the proper process controls, measures can be implemented to contain potential risk. In high reliability applications data logging and traceability is available for each die bonded, but this data is also generated for all standard commercial applications. Thermocompression bonding is perceived as a back end assembly process but the level of process control can be similar to that used in front end semiconductor manufacturing. This paper will review methods available for process control in advanced TCB processes suitable for single die, 2.5D and 3D packages.

An expert system is a computer software technology developed from artificial intelligence research. It may be used for intelligent manufacturing process control and, when properly designed, has the capability to imitate human behavior. An expert system’s value is to assist a human in executing realtime process control decisions in a complex system. This article provides an introduction to the concepts of intelligent process control and includes a process control scenario applying an expert system as well as statistical and optimization technologies. A simple guide of how to get started and a description of expert system tools are also presented.

Quality in laboratory medicine: 50 years on

Microcomputer Control of a Printing Ink Plant

Design by Reification: Process-Centric and System-Centric Models in Process Control

In the new stage for intelligent manufacturing of traditional Chinese medicine(TCM) from pilot demonstration to in-depth application and comprehensive promotion, how to raise the degree of intelligence for the process quality control system has become the bottleneck of the development of TCM production process control technology. This article has sorted out 226 TCM intelligent manufacturing projects that have been approved by the national and provincial governments since the implementation of the "Made in China 2025" plan and 145 related pharmaceutical enterprises. Then, the patents applied by these pharmaceutical enterprises were thoroughly retrieved, and 135 patents in terms of intelligent quality control technology in the production process were found. The technical details about intelligent quality control at both the unit levels such as cultivation, processing of crude herbs, preparation pretreatment, pharmaceutical preparations, and the production workshop level were reviewed from three aspects, i.e., intelligent quality sensing, intelligent process cognition, and intelligent process control. The results showed that intelligent quality control technologies have been preliminarily applied to the whole process of TCM production. The intelligence control of the extraction and concentration processes and the intelligent sensing of critical quality attributes are currently the focus of pharmaceutical enterprises. However, there is a lack of process cognitive patent technology for the TCM manufacturing process, which fails to meet the requirements of closed-loop integration of intelligent sensing and intelligent control technologies. It is suggested that in the future, with the help of artificial intelligence and machine learning methods, the process cognitive bottleneck of TCM production can be overcome, and the holistic quality formation mechanisms of TCM products can be elucidated. Moreover, key technologies for system integration and intelligent equipment are expected to be innovated and accelerated to enhance the quality uniformity and manufacturing reliability of TCM.

The integration of Artificial Intelligence (AI) and Robotic Process Automation (RPA) within Pega’s Intelligent Process Automation (IPA) framework is fundamentally transforming enterprise workflow management. Traditional RPA, while effective in automating repetitive, rule-based tasks, lacks the adaptability and cognitive capabilities required for handling dynamic business processes. AI-enhanced RPA, on the other hand, leverages machine learning (ML), natural language processing (NLP), predictive analytics, and decision-making algorithms to enable self-learning automation systems that optimize workflows, reduce errors, and improve operational efficiency. This study conducts a comparative analysis between traditional RPA and AI-powered RPA within the Pega ecosystem, focusing on key performance indicators (KPIs) such as process execution time, accuracy, cost-effectiveness, scalability, and adaptability. By evaluating empirical data from real-world implementations, this research identifies the tangible benefits of AI-enhanced RPA in automating complex business operations across industries such as finance, healthcare, and e-commerce. The comparative assessment is structured around efficiency gains, error reduction, financial viability, and scalability, providing quantifiable insights into the transformative potential of AI-driven process automation. Using real-world case studies and industry benchmarks, this study demonstrates how AI-enabled automation in Pega improves workflow orchestration, predictive decision-making, and end-to-end automation of critical business functions. AI-powered bots can analyze data, predict process bottlenecks, automate exception handling, and enhance customer interactions, thereby surpassing the limitations of traditional RPA. The findings from this research emphasize the strategic advantages of AI-enhanced RPA in digital transformation efforts. Organizations that integrate AI-powered IPA within their automation strategies gain a competitive edge by achieving greater operational efficiency, reducing costs, and enabling scalable, intelligent automation solutions that adapt to changing business needs. This paper provides actionable recommendations for enterprises looking to leverage AI in Pega-driven automation frameworks, ensuring a seamless transition from rule-based automation to intelligent, self-optimizing workflows. Ultimately, the study concludes that AI-driven RPA in Pega is not just an incremental improvement over traditional RPA but represents a paradigm shift toward autonomous and cognitive automation, setting a new standard for enterprise-level process management.

Nanoimprint Lithography (NIL) has demonstrated its value in manufacturing nanostructures with extremely tight tolerances for the optics industry. Doing NIL process on single samples in the lab has been displayed for a long time in industry and academia. Scaling up from one good sample to full wafer and wafer to wafer uniformity for many wafers in a row, has some obstacles to overcome. In this paper, we discuss the process repeatability of manufacturing nanostructures starting from basic line/space through various other periodic and meta structures. Demonstrating the manufacturing capability using NIL for pattering also includes the other related process steps such as thin films and etching. We discuss key metrology, process control characterization, and process stability which include thin film RI uniformity, master to print replication uniformity, and post etch structural critical dimensions uniformity (CDU). Our NIL process has achieved 1.0nm σ on line widths at 45nm. This precision of replication includes all variation introduced from multiple stamps, wafer to wafer prints, and multiple sites within the wafer. This level of process control at the masking layer needs to be maintained for the finished structure.

Bubble detection is a challenging problem in automatic process control in the power and energy industry, medical and pharmaceutical industry and many other fields. Computer vision methods applications for bubble detection and measurement is the principal step of robust bubbles monitoring systems development. In various applications the input image may include a diverse and image background, especially in different environments. This paper presents a new and effective bubble detection approach. The main steps of this proposed approach are as follows: image preprocessing, background subtraction, and contour detection. The graph cut algorithm is used for image segmentation. The Haar wavelet transform is applied to collect bubble component points. The developed approach is evaluated based on the real data set.

In the construction and development of modern society, the traditional artificial production line operation mode, has been unable to meet the needs of enterprise development, research scholars in the study of the Internet of things as the core of the intelligent production line process control system. The whole system design not only references a lot of advanced technology theory, but also has remote control and automatic assembly and other unique functions. On the basis of understanding the theory of Internet of Things technology, this paper mainly discusses the components of intelligent production line process control system, combined with practical cases to study the system application effect. The final results show that the internal functions of the system are in line with the current production and transportation needs of enterprises, and can solve the problems existing in the traditional production line process control.

A thin film measuring system for the online monitoring of semiconductor films during processing with a radio frequency plasma is described. The method is based upon the use of chromatic processing of polychromatic interference signals produced by the thin semiconductor film when illuminated with white light. A description is given of the interpretation of the polychromatic signals produced by interference effects and some typical experimental results are presented to illustrate the capabilities of the technique. The approach utilizes cost effective instrumentation with acceptable time response for real time process control via optical fibre transmission. It is currently being incorporated as part of an intelligent knowledge based system for such process control.

Intelligent Components for Quality Control in Manufacturing

Computer control of fermentation processes

As we all know, big data and intelligent computing have been widely used in information science, life science, computer science and intelligent control, intelligent robots, vehicle networking, space technology, marine development and other fields, especially in the future life science and medical field. This paper is based on a new digital optimal decision-making idea of intelligent signal processing and control. On the basis of extending non-digital signal processing and information prediction to signal processing and information prediction in the digital generalized information measurement space, and making full use of optimization algorithms such as digital filtering and frequency spectrum analysis, this article discusses in depth the intelligent optimal control of a class of non-stationary random processes in artificial intelligence, biomedicine, radio communication, automatic control and other scientific and technological fields. It provides a new theory, a new approach and an effective and reliable new method for the mathematical modeling and optimal control of such processes, and has broad application prospects and value.

- Marathi is one of the popular oldest languages of India and possesses the greater syntactic complexity. Nouns, verbs and compound words of this language have very clear and simple rules that make the learning of the language an easy task for anybody. However, a more effective tool is required to consolidate the particularity of the higher-level grammar, especially the regional dialects. In the last few years there has been a lot of work done in the era of Artificial Intelligence (AI) and Natural Language Processing (NLP) where various languages related complicated tasks have been made easy. Pre-trained generative AI models like BERT (Bidirectional Encoder Representations from Transformers), GPT-3 (Generative Pre-trained Transformer), T5 (Text-to-Text Transfer Transformer) and many others hold much promise in different language uses like text generation, translation, and grammar checks. These models can formulate, interpreting and modifying any linguistic behaviour that might be useful in handling other challenges of Marathi language such as noun declensions and verb conjugations. AI model built from the transformer architecture can be adapted to handle several of linguistic problems in Marathi language that involves syntax analysis and error detection. This paper presents a generative AI model for learning the Marathi grammar which is further categorized into two parts. The first part of the study is devoted to the elementary grammar training, and the second part is dedicated to the intermediate and advanced levels. Using generative AI, this current model gives higher accuracy than rule-based system and presents effective ideas towards modern grammar. Also, this model became improving tools in computational linguistics for regional languages and for promoting language education. Keywords: Marathi language, generative AI, BERT, GPT-3, T5, computational linguistics