The Antarctic Multiband Infrared Camera (AMICA) will be the first European instrument operating automatically between the K and Q infrared bands on the Antarctic Plateau (Straniero et al. 2007). It will be mounted on the International Robotic Antarctic Infrared Telescope (IRAIT; Tosti et al. 2006) to perform photometric observations of southern sky sources and provide fundamental information on the properties of Dome C, widely considered one of the best terrestrial sites for infrared observations (Burton et al. 2005). The development of AMICA takes up the great challenge of realizing a compact and reliable system, which includes a large number of functionalities, from environment monitoring to camera handling to fast data acquisition and processing (Di Rico et al. 2006). According to the requirement analysis and specification describing the overall functionalities that the system has to provide, an advanced control system has been designed and is currently under development. It is designed to fully meet the numerous and critical requirements arising from both scientific purposes and peculiar operating constraints. Because of the heterogeneity of the system, the resulting degree of complexity is very high. To avoid possible confusion levels, which can result in clashing operations, a well-structured and modularized software design has been carried out. It is ensured by an accurately defined conceptualization of its multithreading architecture, based on the object-oriented approach, favoring the decomposition of the control software in a number of specialized subpackages and providing a more intuitive description of subsystem roles and interactions. Two applications constitute the software control. The AMICA Server Application is the main control software. It ensures the elaboration of the scheduled operations and the correct subsystems configuration. The Detector Control Application is an independent process acting in turn as a server. It is responsible for the configuration of the front-end electronics and for the control of the acquisition process. The successful operation of such a system strictly depends on the synchronization of its threads and on the complete execution of each task, avoiding any possible process hanging or system crash. All solutions adopted for this automatic control system have been carefully thought out in the face of the peculiar site characteristics and local limited facilities. Moreover, an independent low-level environmental control system ensures the electronic equipment safety against dangerous thermal stresses, even if software applications are not running. This avoids possible temperature cycling, moisture condensation, and overheating, which can dramatically reduce the reliability of the system, while the high-level control is strictly required for the camera configuration during ordinary operations. The reliability of this robotic system is the most ambitious challenge. It is the result of many evaluations about all possible incoming scenarios. Every interaction among detailed subsystems has been accurately considered to prevent the generation of any unexpected event. The AMICA control system is provided with innovative solutions against dangerous risks for instrumentation, allowing the performance of robotic operations without human assistance, even in the critical conditions of the Antarctic Plateau. Its hardware and software architecture is designed to be robust and reliable, according to ambient constraints, with the additional aim of achieving the characterization of a standard for Antarctic instrumentation.