Validation tests under different driving and environment conditions
                
              
            
          
            
              
                
                  
                  The framework must provide tools to prepare the datasets for the validation tests in a semi-automatic way.
                
              
            
          
            
              
                
                  
                  Information about all objects detected by the system during the execution of a validation test must be recorded (in a file) for further analysis.
                
              
            
          
            
              
                
                  
                  Accuracy metrics must be calculated for each test execution, comparing the test execution results and the expected results, defined by the ground truth for each journey.