MODIFIED SETUP Sample Clauses

MODIFIED SETUP. Additional datasets collected after the system was modified and during favorable weather conditions were also selected and analyzed. This analysis for the modified system setup was based on 6 days between November 17, 2012, and January 14, 2013, in favorable weather conditions. Selected data from the modified setup covered a similar sample size in terms of the number of hours and traffic volume as those presented for the initial setup. The summary of the analysis results in terms of frequency of errors by day is shown in Table 3. About 0.96% of the total number of calls generated by the system, including both radar units and lanes, were false calls. The majority of those false calls (0.59% of the total calls) were created when the gates were moving, either being lowered or raised because of train arrivals and departures. This change is significant, compared with the initial setup, for which this type of false calls had a much lower frequency (0.03% of the total). By contrast, the frequency of false calls caused by bicycles and motorcycles was lower in the modified setup, compared with the initial setup. This change could be attributed to the lower use of these two modes during the cold months. The frequency of pedestrian false calls remained at a similar level. NORMAL WEATHER - Modified Setup (Winter 2012-2013) Date Activations Trains False calls (including bicycles and pedestrians that radars detected, but loops did not) Missed Calls Stuck-on Calls Not detected by subject radar in any lane Not detected by any radar in any lane Wavetronix Loop Bicycles Pedestrians No object Gates Moving TOTAL Nov 17, 2012 (24 hrs) 6464 6098 97 11 37 5 68 121 2 0 0 Dec 14, 2012 (24 hrs) 7926 7354 143 2 11 8 38 59 1 0 0 Jan 14, 2013 (24 hrs) 6738 6304 124 0 8 0 25 33 7 0 0 * No cases of motorcycles detected by the radar but not by the loops were found. The total number of motorcycles using the crossing and detected by the two systems is unknown. Regarding missed calls, the frequency of vehicles not detected was slightly higher in the modified setup (0.09%), compared with the initial setup (0.07%). Also, in the selected datasets, there were no cases of a vehicle not being detected by the two radar units. Thus redundancy in the radar detectors worked appropriately. Stuck-on calls and dropped calls were not found in the modified setup during favorable weather conditions. The performance of the system with the modified setup, by radar unit and lane, is shown in Table 4. In gener...