Managing Infeasebilities in a Hub and Spoke model

My independent study at CMU during summer semester was related with finding feasible shipment delivery schedule and plan for infeasibilities in a hub and spoke model. I was very much intrigued to know how huge networks operate in a hub and spoke model where a small delay at any stage can have a rippled affect on the service levels. It was an instinctive decision to study the operational model of logistics giants such as UPS, Fed-ex and DHL and find out how these organizations operate and what makes these companies the best in their field.

Well, I decided to read about UPS and landed up watching the video of their World port located at Louisville, KY, a massive infrastructure that acts as the worldwide hub of UPS. For a company that delivers over 4 billion packages every year in more than 200 countries, a HUGE infrastructure that comprises of more than 200 planes and over 9000 vehicles acts as its backbone.  The video talked about every nook and corner of the world port hub and showcased how UPS has fine-tuned every bit of its internal processes so as to be efficient operationally and provide excellent service levels to its customers. Operationally the whole model looked highly efficient in nature, and the level of automation achieved within the world port would blow-off any sane mind. 

But a model of this nature, which has a lot of dependencies involves external agents that can easily make the operations go hay-wire. Occurrences such as bad-weather, peaks in air-traffic, natural calamity or even a technical snag can affect the operations adversely. An interesting question, which comes to my mind, is that with such hub and spoke models, what kind of algorithms are applied by these companies, which  -

        a) Guarantee optimum utilization of resources, such as vehicles, planes and rail networks etc.

             b) Ensure the service levels promised to customers are met

      For example, a UPS flight carrying shipments from London to Louisville would be carrying shipments destined to various parts of US. Some of them might be priority shipments, where as some might require further forwarding through road and some via air. A small delay in the flight schedule or a delay at the hub due to a technical snag will have a direct affect on the actual delivery time of the shipments, and at the same time a n alternative solution might not be cost-effective. Another example can be how a vehicle,on the road , which is supposed to pass on certain shipments to another vehicle, can go down on the way and thus, affect the delivery of those shipments. 

      How companies having such huge networks plan for such infeasibilities, or do they plan for such infeasibilities and then come up with their promised service levels. What are the kinds of damage control mechanisms applied and how are customers given a transparent picture of such slippages. Such occurrences pose questions on the service levels and an interesting way to go about these problems is to understand how technology can be leveraged to plan and overcome such infeasibilities?

      I am attaching the link of the video along with this blog post,(this is a 3-part video)

http://www.youtube.com/watch?v=KWAwL3Rlbjo.

Looking forward to some interesting comments.