Issue link: http://www.e-digitaleditions.com/i/579724
16 OCTOBER2015 Inhalation Design of an inhaler-specific robotic system for laboratory use in early development Introduction The complexity of current inhalation devices requires thorough testing of all aspects of device reliability. Classic manual testing constitutes an important part of the process, while for the sake of completeness, human factors engineering and usability studies address the variability introduced by the human user. However, as far as reproducibility is concerned, automation can provide strictly predefined handling at high throughput, which may be beneficial in obtaining the low variability required for the predic- tion of reliability, in view of expected failures in the low parts per million (ppm) range. However, depend- ing on the specific situation, poorly-designed auto- mated tests may miss important failure modes simply because of the deterministic way of testing. This article addresses the challenges of automated testing in early development, where two additional concerns arise: 1) the need for fast adaptation of test equipment when shape factors change, which can happen often in early development and 2) the need for high throughput and reliable test results, which are ensured by lean quali- fication and validation procedures. In this article, robotic testing is presented as an additional voluntary effort during development of an innovative multi-dose inhaler, yet the compendial, conventional, manual release tests were still in place and carried out in compli- ance with all applicable regulations. The inhaler evalu- ated by this robotic system delivers liquid droplets from a non-pressurized reservoir. A spring-operated piston pump inside the inhaler meters the dose and generates the necessary pressure for atomization. Therefore, some aspects of the system are specific to this type of inhaler, However, many aspects of the robotic system will be applicable to other aerosol delivery devices. General considerations for automated versus manual testing Automated testing represents a considerable investment, not only in hardware but in labor. The expertise required to design an automated system calls for both technical and regulatory competence. Lead time is required if a new automation project is to be launched. Often the design of dedicated robotic test modules is time-con- suming while qualification and validation activities can prohibit efficient use of automated test systems during early development. In comparison, manual tests may be started faster and can be more flexible. However, that quick-start advantage must be balanced with the unique reproducibility and speed of robotic tests, as well as the reduction in human labor, once a robotic system has been established. If time to market is the highest priority then manual tests are likely to be selected. Yet in that sit- uation, testing specifications must consider future automation. Experience and an interdisciplinary team are required to enable optional automation at a later point in time. For example, if a delivered dose measure- ment were to be automated, the consumption of solvent might be lower if performed manually by experienced personnel. The (common) testing specification must be written with future automation in mind, in a way that man and machine can run the test. This exchangeability is the ultimate goal of automation. If an increased num- ber of similar tests motivates large-scale automation for Herbert Wachtel, PhD Boehringer Ingelheim Pharma GmbH & Co. KG Challenges and opportunities to advance product development A set of trays giving access to 60 inhalers. In total, five similar racks can be program-selected, providing storage capacity for 300 devices.