This section shows measured results in an urban environment where the primary pollution originated mainly from the city traffic i.e. from the internal combustion engine. Major intersections were selected with high measured concentrations of pollutants coming from the motor vehicles exhaust pipes. In the vicinity of the intersection there were no other sources of pollution that would dominantly influence the measured values. The concentrations of the following gases were measured in the air: sulfur dioxide (SO2), carbon monoxide (CO), carbon dioxide (CO2), particulate matter (PM10), relative humidity (RH) and air temperature. The monitoring was carried out over a long period of time, with a sampling interval of one hour. Program for the reception of data from monitoring stations and statistical analysis of the measured values was implemented in LabVIEW software package on a PC under Windows operating system. Data from monitoring stations are transferred to the server via the Modbus protocol, which enables fast and reliable transmission of data. Fig. 5 shows the front panel of HMI (Human Machine Interface) displayed on the server computer. It displays the measured values of environmental parameters in numerical and graphical form. The graph shows four curves. The color red is used for the measured concentration of SO2 gas in the last 18 h. Current value of SO2 gas is also there in numerical form and at the scale display. Black color displays the numerically computed prediction of SO2 gas concentration in the last 18 h as well as the estimated value of the gas concentration over the next 6 h. Green and blue colors are used to graphically depict numerically calculated, expected, lower and upper limits for a period of 24 h. Estimating gas concentration in the next period is done using the ARMA model, which is implemented in the LabVIEW program. Since the ARMA prediction relies on the previously obtained values it is possible to calculate arithmetic mean (μ) and standard deviation (σ ) of the measurements. The upper and lower limits are now determined as View the MathML source with probability of 95.5% (2σ). In this way, if unexpected high concentrations of pollutants in the air occur, users would be able to take certain actions. These activities consist of avoiding these locations as well as using protective equipment when necessary. Since it is a statistical method it is not possible to give exact measurement errors because it depends on each newly arrived data and therefore it is much better to give the lower and upper limits which will move around the prediction values. ARMA model relies solely on a single parameter that is observed while other statistical models, such as mathematical correlations and regressions rely on the mutual dependence of different types of gases. Certainly, it would be very useful to improve the method using concentrations of other air pollutants, but it requires a long-term research and complex mathematical calculations.