Paper Title
Intelligent Quality Compliance System for Textile and Apparel Businesses

An Intelligent Quality Compliance System has been developedto generate regulatory/health/environmental guidance reports for textiles and apparel companies. Subsequently, the system will be augmented to incorporate additional consumer goods. The apparel category provides for a diverse set of mandatory regulations and some voluntary standards. We address mandatory requirements such as CPSIA, FTC for Care and Textile labeling, in addition to AATCC requirements for colorfastness, formaldehyde, etc. Databases from the International Agency for Research on Cancer (IARC), the US National Toxicology Program (NTP) are to be incorporated,in conjunction with computational intelligence/chemistry, to identify potential toxins or carcinogens present in the industrial process or the final product, thus alerting manufactures and consumers through a user friendly interface. There exists an industry need to further develop an IT solution to provide an automated process for the identification of global product safety requirements based on pre-defined product criteria with a tool to enhance the efficiency and accuracy of determining reasonable testing programs to insure safety of products being placed in the global market place. The escalation and complexity of regulations and testing requirements, as well as their access is of vital concern to decision makers in the textile/apparel industries. Currently the process of identifying required testing regulations is accomplished by hand, and this processis time consuming and prone to oversights. In addition, the current set of recommendations does not cross-reference International Agency for Research on Cancer (IARC) and the US National Toxicology Program (NTP) databases. The unique advantage of the proposed technology lies in harvesting hidden relationships through the computational intelligence/chemistry power of artificial neural networks, genetic algorithms, and fuzzy systems. The proposed tool provides the end user with unique knowledge to satisfy regulatory requirements on one hand, and to differentiate and add value to the consumer products on the other hand. The driving force behind this differentiation lies in utilization of computational intelligence approaches, such as artificial neural networks, genetic algorithms, and fuzzy systems. Those approaches are much more adequate for dealing with uncertainty and complexity of factors involved. Using the proposed expert system will enable decision makers to retrieve all relevant documents related to regulatory/testing information, AATCCrequirements, data on potential health/environmental concerns, as well as provide suggestions on improving materials and development practices.The system will present its findings in a coherent and cohesive way to the user, thus enabling informed decision making. The scope of this research is to create a prototype expert system to integrate industry regulations as well as to address consumer concerns, expressing them as expert rules, and subsequently translating and displaying them through a friendly user interface. A software package that integrates and summarizes required testing/regulatory information, ATTCC requirements, data on potential health/environmental concerns, as well as recommending improved, cost-effective materials/processing procedures, and presents the results in a coherent and cohesive way to the user, all in one place, will be invaluable to textile/apparel companies. We have worked closely with industry to assure the relevance, accessibility, and potential impact to a large segment of the textile/apparel industries. Incorporating the identification of toxic agents in consumer textiles/apparel and their processing makes this expert system unique. The innovation lies in integrating computational intelligence and computational chemistry to identify potential toxins and carcinogenic agents and alert consumers/manufacturers.These approaches are much more robust in dealing with uncertainty and complexity of factors involved in potential toxin formations than any statistical analysis and tools currently utilized in chemical compound recognition and classification. In the future we envision that the proposed expert system will also provide an indispensable tool for clinicians to identify potential cancer type in patients who have a certain genetic makeup by running corresponding compound matches. The proposed expert system requires using existing tools: CODESSA and Neural Works to generate entries into the system. Index Terms- Intelligent System, Quality Compliance, Textile and Apparel Businesses.