Development of a Method to Evaluate Product Defects in Trucks - Creation of a method to simplify recall decision process of a defect that can lead to personal injuries

Abstract

The number of vehicle recalls due to product defects has increased during the last years. A large number of recalls is expensive for a company, both by direct costs as well as damaged reputation. If a defect of a part is detected, an analysis is performed to investigate if the defect can cause personal injuries, as well as how likely it is for the defect to occur. The objective of this thesis was to create a general method to evaluate defects that potentially can cause harm, as well as rate them regarding severity in order to support the recall process. The method is based on different types of existing hazard analyses. The hazard analyses most used are FMEA and ISO 26262, where it is found that each defect should be evaluated with regards to severity, probability of exposure and controllability. Severity rating tables are created by studying data from STRADA, injuries in maintenance and efficiency of safety systems, among other things. Probability of exposure rating tables are created by studying, for example, usage specifications of sold trucks and Swedish road data, as well as maintenance routines and usage of safety systems. The controllability rating tables are created by studying similar evaluations in ISO 26262 and the occurrence of the possible defects. All hazardous scenarios (i.e. all possible harmful situations that can take place if a defect occurs) are identified by studying internal defect data as well as external recall data. Possible crash scenarios resulting from the hazardous scenarios are found, and the corresponding rating for severity, exposure and controllability is identified for each scenario. The ratings are combined into an ASIL rating, which clarifies if the defect is safety related (i.e. can cause any personal injuries), as well as a rating how severe the defect is. The highest ASIL is identified for each hazardous scenario and occurrence, and these are gathered in tables sorted depending on the effect for the customer. When using the method, the analyst has to identify the hazardous scenario the defect can result in and the occurrence of the defect (e.g. sudden, previous notice). After this, the worst case ASIL is found from the applicable table. If the current hazardous scenario does not exist, the case have to be analyzed using the tables for severity, probability of exposure and controllability. In conclusion, the method simplifies the evaluation of the severity of a defect, and creates a rating which can be used to support the recall decision. However, the method has to be verified further on new potential safety hazardous cases, as well as on different analysis engineers to ensure repeatability.