Reducing Errors in Development Trials Using Technology Transfer in Contract Chemical Manufacturing Open Access
Downloadable ContentDownload PDF
Contract chemical manufacturers spend resources on development trials that have errors due to incomplete knowledge transfer from the customer. These errors waste engineering resources and can result in lost business. A technology transfer program was proposed as a solution to reduce the number of errors. To determine the contents of the program, data was collected from a contract chemical manufacturer. A model was created using partial least squares regression to predict the number of errors based on trial characteristics and technology transfer activities. Two-sample Poisson rate testing was used to test the manufacturer’s existing technology transfer approach. Lastly, case studies were used to illustrate how the proposed program would reduce transfer errors.The partial least squares regression model showed that plant scale trials, mature projects, and less mature products tend to have fewer errors. The model showed that technology transfer activities led to more errors, contrary to expectations. This was attributed to the manufacturer selectively using the additional activities for the more complex and difficult projects. The two-sample testing showed that only some of the transfer forms were significant in reducing errors. The recommendations for the transfer program were: 1) Have a technology transfer program in place with structured and unstructured transfer methods with support from management. 2) The assigned engineer should witness a batch being produced at the customer’s facility (preferred) or in the manufacturer’s lab. 3) The customer should supply a sample of the product, so the manufacturer’s Quality Control lab can run analytical testing to verify the capability of the equipment and technicians and to confirm accurate results. 4) The customer should have technical staff monitor the development trial alongside the manufacturer’s engineer. The case studies showed that these recommendations were supported by the literature and could have reduced errors in each of the cases.