Alberto Guiggiani, University of Firenze, Italy - Department of Systems and Informatics
Michele Basso, University of Firenze, Italy - Department of Systems and Informatics
Massimo Vassalli, Institute of Biophysics, National Research Council - Genova, Italy
Francesco Difato, Italian Institute of Technology (IIT), Department of Neuroscience and Brain Technologies (NBT) - Genova, Italy

When building a real-time machine, open-source software can constitute a suitable alternative to commercial solutions in terms of functionality, performance and costs. On the other hand, the major drawbacks of this choice are the long amount of time and the advanced computer skills needed to make all the pieces work together. These steps represent an obstacle that might discourage less-experienced people, like students for example, to enter the world of Linux-based real-time applications.

Our aim is to provide the tools and documentation needed to easily set up a working real-time machine for signal acquisition and processing. In order to achieve this, we started by collecting all the pieces required to configure a complete software chain: an Ubuntu operating system with RTAI modified kernel; Comedi to interface with DAQ boards; Scicoslab with RTAI-Lib palette to generate and build the real-time target and, finally, QRTAI-Lab and RTAI-XML to monitor the target locally or remotely, respectively. The next step was to edit source codes with the objective to avoid conflicts and compile errors. Lastly, we packed everything in the so-called Real-time Suite alongside with documentation with simple step-by-step instructions and examples. In addition to that, we configured the Suite on a virtual machine that works out-of-the-box, useful for testing purposes.

A Real Time Suite based system has been installed to develop a feedback control for nanometer-precision specimen tracking in an optical tweezer system with piezoelectric actuators.

This real application will be presented, and the achieved performance will be discussed, including latency and jitter across the control loop as well as supported sampling rates related to the number of channel used.

With Real-time Suite, we have configured a ready-to-use software package useful for researchers/users approaching real-time applications for the first time. Indeed, they can build their own machine with only a few hours of work and at the cost of a personal computer with a supported data acquisition board.