The objective of this thesis is the implementation of advanced mathematical methods for inverse problems on real-time control computers. The required code is generated automatically from an abstract model using automated code generation. Modern cyber-physical systems have to perform complex computations and versatile communication tasks simultaneously. The software development process can be improved utilizing code generation and a real-time operating system on the target controller. The exemplary application chosen to demonstrate the methods in this work is the inverse problem associated with the reconstruction of deformation-curves from the data acquired from an array of inclinometers, i.e. from local gradient measurements. In the laboratory the sensors are mounted on a beam, whose deflection is computed with a recently developed solver for ordinary differential equations. The algorithm is implemented on an embedded prototype system running under real-time Linux. Code is generated automatically from Matlab/Simulink. Communication with the sensors is performed over a RS-485-based field-bus. Ethernet UDP is applied to transmit data between tasks on the target system and to the operator interface running on a PC. The correct functionality of the embedded system, and with this, the complete development workflow, is verified using reference measurements in the laboratory.