An analysis of the operational performance of KLM´s Baggage Turnaround Services.

Abstract

Since 1919 KLM operates from its Hub at Amsterdam Airport Schiphol (AAS). Besides the air transport of passengers and cargo, KLM performs ground processes at AAS. Of which one is the handling of the passenger baggage between the airplanes and the baggage handling systems of AAS. Every year KLM’s division of Baggage Turnaround Services (BTS) handles around 35 million pieces of baggage of 130.000 flights at AAS. Including the baggage from Sky Team airlines and other client airlines of KLM BTS. The aim of BTS is to provide a safe and reliable service for KLM and its client airlines. The reliability of this service can have consequences on the customer satisfaction of the passengers, the departure punctuality of the outbound flights and the costs for KLM in redirecting mishandled bags. In 2010 KLM BTS was responsible for around 450.000 mishandled bags which costs KLM € 50 million only for the redirecting of the baggage. A mishandled bag (MHB) is a bag that does not arrive on time on the final destination of the passenger. The consequences in terms of customer satisfaction and departure delays are not even included in this amount. Benchmarks of the Association of European Airlines (AEA) show that KLM`s European competitors perform (transfer airlines: Air France, British Airways and Lufthansa) structurally better than KLM based on the number of mishandled bags per 1000 passengers (=Irregularity rate). KLM has the ambition to become the most reliable European airline in terms of baggage handling within two years. In order to realize this KLM BTS needs to better understand how its (non-)performance is being caused. This research will provide these insights for KLM by answering the following research questions: What are the main influences on the operational performance of KLM`s Baggage Turnaround Services? To answer this research question a research approach that combines different qualitative and quantitative methods is designed. The approach consists of an actor analysis and a process analysis which generate potential influences divided into internal and external stakeholders of KLM. These potential influences are subsequently tested with statistical and data mining methods. The data mining methods consist of classification trees and neural networks. The operational performance of BTS is being defined as the Irregularity rate (Irrate in MHB`s/ 1000 passengers) to reduce the scope of the research and because of the limited data availability. As 98% of KLM`s Irrate score is caused by baggage that has a transfer at AAS, the research is concentrated on the transfer baggage flow. The qualitative research has generated first insights in how the Irrate can be influenced. It shows that the Irrate is influenced by many different factors of which a great part do not lay in the power of BTS. The major influences are classified in four stakeholder groups and some external influences which cannot directly be influenced by any stakeholder. The four stakeholder groups consist of: KLM, AAS, BTS and external stakeholders. These different groups can each in their own way have an impact in the baggage handling performance: 1. KLM influences consist mainly of network related factors as the distribution of flights and connection times but also the defined arrival and departure punctuality targets. KLM’s network targets are conflicting with the operational performance of BTS. Short connections are attractive for the potential passenger but difficult to realize in terms of baggage handling for KLM BTS. The same accounts for the on time departure (D0) target. 2. AAS determines the capacity and functioning of the baggage handling systems but also the distribution of the baggage of all non KLM BTS flights over the day. With respect to the baggage handling performance AAS has the same goal as KLM BTS. Nevertheless AAS also has to cope with the interests of three other ground handlers and more than 100 different airlines. 3. The group of external stakeholders that influences the Irrate consist of the outstation that load the airplanes and also often influence the arrival punctuality, customs that can perform all kind of security checks in the baggage basements and air traffic control who influences the arrival punctuality by assigning landing strips and prioritising certain flights. The load compliance, the choices to scan or check specific flights or the changes in arrival times all influence the difficulty of the task and therefore the performance of BTS. 4. KLM BTS itself can change process handling rules, buy more and or better materials and information systems and train and hire more personnel. The results of the quantitative research are for a large part determined by the data availability of the potential influences described above. First of all a regression analysis has shown the causal relation between the arrival punctuality (AO) and the Irrate score on a daily level. The A0 performance can explain 32% of the variance in the Irrate performance. No further significant relations have been found on a daily aggregation level. The use of the classification trees has emphasized the importance of the connection time of the transfer baggage. Baggage with short connections are far more vulnerable. Furthermore the lead time of the process between the arrival of the flight and inserting the baggage into the baggage handling systems influences the chances of baggage becoming an MHB significantly. The use of the neural networks have confirmed some of the insights obtained from the use of the classification trees and most importantly it has shown the difficulty in predicting MHB`s with the current available dataset. Neural network could only predict 47% of the MHB`s. This confirms the lack of useful and reliable data of the sub processes performed by BTS. Most important conclusions for KLM are that many of the variations in their performance are based on the performances of non BTS divisions or organizations. Where possible the impact of the variance in these factors on the Irrate or MHB percentages is provided. Furthermore a list of opportunities to minimise these influences is composed. Most important opportunities for KLM BTS are to focus more on the inbound load (collaboration with outstations), set specific arrival punctuality targets for inbound flights with many short connections (collaboration between divisions within KLM), exchange more data with AAS(collaboration with AAS), and focus on KLC outbound and ICA inbound processes (BTS) as these are more vulnerable for MHB`s. The research approach has confirmed some of the gut feelings that existed within KLM BTS and it has provided KLM with some new insights of vulnerable processes in their baggage handling. Nevertheless the full potential of the data mining methods can only be reached if more lead times of the different sub processes of BTS are being gathered and if the dataset is fully reliable in terms of BASS scan moments. This could generate models with a higher prediction accuracy in terms of MHB`s and it would provide more insights about the critical sub processes within BTS.