Industrial statistics is an important part of the management system in any industry that strives to continuously improve quality and increase productivity and efficiency. That system covers supply chain management, production design and prototyping, production process and marketing. Industrial statisticians, industrial engineers and industrial leaders should work together hand in hand, in the same language, to ensure that the process and products are as expected. The system itself is never complete. Thus, the usefulness, manageability and reliability of all statistical models used in the system are to be considered as first priority, but those skills are not sufficient. Industrial statisticians should also, of course, be able to come and go between the two poles: statistics and industry. This requirement needs a good understanding about the culture of these poles and how to conduct a mutual symbiosis. One of the principal bridges between these cultures is statistical process control (SPC). This paper is to show that modern industry cannot escape from SPC, especially in a multivariate setting. This setting, which characterizes modern industry, consists of two philosophical problems: how to order data and how to measure process variability. Our recent research results sponsored by the Government of Malaysia will be presented to illustrate the challenging statistical problems in modern industry.