Radiation detectors featuring long term stability, self-sustained operation and low power consumption are crucial for longterm environmental monitoring (e.g. nuclear waste disposals and mining activities) and provide enhanced applications of nuclear fingerprinting e.g. in farming and geological surveying. INCAS 3 is developing a compact modular system consisting of four functional modules, namely analogue conditioning and signal digitalization, dead-time-free real-time signal processing, embedded high level analysis of the processed signal, and wireless communication. The modules are organized such that they can be interchanged and modified independently. For the input module one can choose an ADC sampling frequency to be either 100 MHz with 14 bit precision or 1 GHz with reduced precision (10 bit). The main focus of the signal processing section, based on an FPGA, is on providing dead-time-free signal handling in real time. Other useful features such as base line correction, pulse shape analysis (energy, decay and arrival time) are being developed as (VHDL) library functions. Additional modules, e.g. anomaly detection in the incoming signal, pile-up correction if operated at high rates and advanced signal shape processing, can be included in the processing if required and can be applied to autonomously generate the information necessary to control the sensor parameters and stabilize energy spectra and sensitivity. At present we operate the system in conjunction with inorganic scintillators (NaI, CsI) read out by a photomultiplier in order to provide a system capable of long term quantification of nuclear contaminations in natural environments. The underlying technology is based on detecting natural or anthropogenic gamma radiation and generating corresponding energy spectra in real time. The generated spectra are analyzed either in a standard way by any suitable desktop software in a lab or, as it is described in this work, by the ENSA (Embedded Nuclear Spectra Analyzer) board. ENSA returns the nuclides activity concentrations of the gamma source under examination, in real time and in the field. The analysis, aimed at nuclear fingerprinting, is based on a Full Spectrum Analysis method developed by Medusa Systems BV. This analysis is usually performed for the natural occurring radionuclides: Potassium ( 40 K), Thorium ( 232 Th) and Uranium ( 238 U), but can potentially take any other nuclide into account. By embedding the data analysis in the sensor device, one creates a fully integrated system. When equipped with a local user interface, the device becomes a portable system that provides highly accurate feedback in the field. With wireless communication, a self-sustained system featuring long term stable operation for environmental monitoring is created.