The whole system is simulated using an Euler integration method with a time step of 1 ms. two sensor inputs & two mortor output 2-D unlimited arena The neural network consists of 6 nodes, fully-connected except for self-connections neurons are either excigtatory or inhibitory and this is set genetically The Poisson spike trains coming from the two sensors are fed into neuron n2 and n3, respectively. Additionally, uniform noise is present in the sensors with range 0.2 -- this results in spikes that fire randomly with very low probability when the sensor is not stimulated. Two mortors control the robot wheels. Each motor is controlled by two neurons, ont that drives it forwards and the other one backwards, using a spike-based leaky integrator. The left motor is controlled by n0 (forward() nnd n4 backward) and the right by n1 and n5. A population of 30 robots is evolved using GA with truncate swelection. Initial weiths are randomly chosen at the start of each evaluation from the interval [0, w_max] <200 Hz -> 200 in 1000 msec 1 spike every 5 msec> Every time a spike arrives to neuron j from an excitatory presynaptic neuron i the excitatory conductonce of j is increased by the current value of the synaptic strenght wij(t). The inhibitory conductance g_in is similary affetctd by spikes coming from inhibitory neurons conductance otherwise decay exponetially