All entries for Monday 26 July 2010
July 26, 2010
Find extreme or good examples of the following practices and justify the reasons for their adoption:
1. Level capacity management
2. Chase capacity management
3. Yield management
4. Queue design
Different business and operations are adapting different capacity managements. Capacity can be defined as an output over a specified period of time. Final designs of capacity management depend on customer demand and nature of operations (Walley 2010).
1. Level capacity management
Slack et al. (2006) describes level capacity management as a process in which processing capacity is set at a uniform level through the planning period, regardless of the fluctuation in demand. In case of non-perishable materials production, if not immediately sold, these can be transferred to finished goods inventory in anticipation of later sales. As captured in Figure 1, setting capacity below the forecast peak demand level will reduce the degree of underutilization, but, in the periods where demand is expected to exceed capacity, customer service may deteriorate.
Figure 1. Level capacity management – absorb fluctuations
Silicon is a metalloid widely distributed in dust & sand. Amongst its main industrial uses are: usage as components in semiconductor devices, but it is mainly used in integrated circuits (microchips). Production of silicon involves several steps in which pure quartz, at a cost around $1.7 cent per kilogram, is transformed into monocrystalline silicon rod worth more than $500 per kilo. Production of silicon is therefore rather costly and involves complicated technological process which incurs high fixed costs. It is not flexible to fluctuate based on demand, manufacturers are aiming to achieve economies of scale. Usage of level capacity management is most suitable approach to production of silicon. Products that are not immediately sold are stored in warehouses. Silicon is a non-toxic material which is fairly easy to store.
Picture 1. Silicon wafle
2. Chase capacity management
As portrayed in Figure 2, chase capacity management is an attempt to match capacity closely to the varying levels of demand (Slack et al. 2006). To achieve this, company has to vary its resources according to demand, which is not suitable for capital-intensive operations. Adjustment of capacity in chase capacity management can be achieved by several methods (e.g. overtime, annualized hours, staff scheduling, etc.).
Figure 2. Chase capacity management
IBM Mainframe customers, which are typically large enterprises (e.g. banks), can already for quite some time use the offering called “On/Off Capacity on Demand”. The idea behind is very simple and provides fast response to customer demand. Customers are purchasing from IBM Mainframe products with originally built-in high capacity, however they can choose what capacity they want to use and respectively pay only for selected level of capacity – customer uses only specified number of processing units while others remain disabled. Should the customer require increase of capacity due to peak of operation, instead of lengthy upgrade process, he can purchase fast access to extra processing unit already available on the machine. There is no physical shipment involved only access code is provided. Once normal workload levels resume, customer can choose to turn off the extra capacity. In this way IBM is matching fluctuation in customer demand, rather than selling level capacity, which is increased in steps without possibility of decrease (Figure 3). In such instance customers were facing underutilization of extra capacity in low demand periods.
Figure 3. Staged level increase capacity in standard mainframe purchase
3. Yield management
In yield management, as described by Slack et al. (2006), service providers always try to use the capacity to its full potential. This is especially used when:
- Capacity is relatively fixed
- Service cannot be stored
- Service is sold in advance
- Making a sale is at relatively low cost
Number of methods is used to ensure operation is maximized (e.g. airlines overbooking).
One of the closest examples for yield management that comes into my mind is a beauty studio. Demand of women for beauty services was, is and probably always will be high. Premises of the studio are typically rather small, capacity is fixed. Rent for the studio depends on the area, however attractive areas have rents rather high and labor is relatively cheap. Studio therefore wants to be utilized to maximum and on top employees are willing to work harder to make more money for extra hours, but also for tips. Beauty treatment is a service that cannot be stored and most of the beauty services can only be performed at that particular place mainly due to equipment used. Customers are required to make an appointment well in advance and since typically there are customers who did not make an appointment trying their luck, in some studios service stuff contacts every customer shortly before the appointment to ensure it is still valid. This depends on the demand levels of course, some studios simply take available customer in case the scheduled one did not show up on time. Yield management is the best method to be used by beauty salons in order to maximize their profit.
4. Queue design
Good capacity plan acknowledges existence of queues and therefore designs the queue. As categorized by Walley (2010) we can allow the queues to form in two different forms:
a) single queue with multi-server design
b) multi-queue with multi-server design
Customer behaviour in queues is also a very relevant factor. How fast can they proceed? How they can get entertained while waiting? These are all the aspects that should be considered when designing a queue.
Since I had the chance to observe different types of queues in different airports in the world, for this exercise I would like to compare three airports and their queue management at the arrival passport control. This is a particularly painful check point, as there is no way for the passenger to turn around and come another time, nor they can choose a different way of exit, while mostly they are anxious and under time pressure to leave the airport as soon as possible.
1. Dubai Airport
As pictured in Figure 4, arrival passport control in Dubai airport is a combination of two queue option designs described by Walley (2010). There is more than one server available per each queue. This design is very effective. Not only customers can easily move fast in the queue and select the shortest one, but speed at check points is higher due to more servers available. On top, there is always airport stuff regulating not only the number of people in queues, but also ensuring that people do not block the process by moving too slow to one of the free servers due to not noticing a free server.
Dubai airport has established also fast track which is called E-Gate. The idea behind is that Dubai residents can purchase a pass which is added on their ID card. Card contains all required data about the resident in electronic form. Passengers scan their ID card at the E-Gate check point and proceed to the exit. Procedure is very fast and helps to off-load the passport control. Frequent travelers appreciate time savings allowed by E-Gate procedure.
Figure 4. Arrival passport control Dubai airport
2. Vienna Airport
Vienna airport passport control upon arrival (Figure 5) is a typical example of multi-queue with multi-server design (Walley 2010). This queue design allows steady pace of the queue. There are queues for EU citizens and non-EU citizens as well, which helps to speed up the process. There is, however no fast track for passengers. Fast track is only available for crew members, airport stuff and disabled passengers. Considering the volumes of passengers arriving to Vienna airport, this design often becomes insufficient. Airport and authorities should consider increasing number of servers as well as fast track for EU-citizens.
Figure 5. Arrival passport control Vienna airport
3. Birmingham airport
Queue design on Birmingham airport (Figure 6) is a typical example of single-queue with multi-server design (Walley 2010). The pace of the queue is rather slow, the arrival hall is small, so the atmosphere is tense. Similar to Vienna airport, fast track is only available for specified personnel. Birmingham airport is a small airport compared to Vienna and Dubai, therefore need for different design or more servers might not be justified.
Figure 6. Arrival passport control Birmingham airport
Hardware – IBM mainframe On/Off Capacity on Demand [Online] (http://www.mainframe-upgrade.com/mainframe-ibm-capacity-on-demand.php) (Accessed 25 July 2010)
Silicon and Steel [Online] (http://www.softmachines.org/wordpress/?p=261) (Accessed 25 July 2010)
Slack, N., Chambers, S., Johnston, R., and Betts, A. (2006). Operations Process Management. London: FT Prentice Hall
Walley, P. (2010). Operations Management. Coventry: Warwick Business School
Wikipedia [Online] (http://en.wikipedia.org/wiki/Silicon) (Accessed 25 July 2010)