Spike Analysis Framework is a new flexible approach for neuronal cell analysis. Advanced engineering and information processing principles will be subject of the. Neuronal cells have the remarkable characteristic to generate electrical impulses for communication. These potential differences are known as action potentials or simply spikes. There are several commercial fields of application for this knowledge: pharmaceutical industry (to reduce the amount of bioassays), in medical science the understanding of neuronal diseases or the applications in robotics and artificial neuronal networks. Communication between neuronal cells is provided. The spikes can differ in amplitude (μVolt range), form and duration (range of 10 kHz). The Multi-Electrode-Array (MEA) system records up to 256 parallel channels. The signal analysis pipeline has to handle a large amount of data in long term recordings. The Spike Analysis Framework solves the task by using relational databases. The advantage is that the amount of recordable data is stored in terabyte range for actual database systems. Only actual processed data chunk must fit into main memory. Parallel analysis and shared analysis over multiple analysis nodes in Ethernet networks are possible. Subject to the analysis is un-mixing the signal mixture from the MEA (known as cocktail party problem in speech analysis) and the classification of specific spike signals and their changes over time. Further a solution for the spike separation and classification is implemented by an embedded acquisition system. To improve spike analysis a combination of algorithms from speech analysis and computer vision is used. Better analysis results in un-mixing and sorting procedures are accomplished, if optical and process control sensors like pH sensitive electrodes and temperature sensors are integrated into the sensor system.