Mgt 3121 Cheat Sheet

Goal of op management: Matching supply and demand while making a profit

Capabilities: the dimensions of the customers utility function a firm is able to satisfy. This leads a company to use
trade offs. Strategic trade offs: a firm must choose a set that excels in one dimension but not others

Pareto Dominated: the firms products and services are inferior to their competitors on all dimensions (inefficient)

Inefficiencies: Waste: all the consumption of inputs and resources that do not add value to the customer. The
customers do not want to pay for waste. Variability: changes in supply or demand over time (Demand – customer
arrivals, customer requests, and customer behaviors. Supply – time to serve a customer, disruptions, defects.
Flexibility: the ability to react to variability). Inflexibility: the inability to quickly and cheaply react to variability

Process metric: something we can measure that informs us about the performance and capability of a process
Inventory: the number of flow units within a process Flow rate: the rate at which flow units travel through a
process Flow time: the time a flow unit spends in a process from start to finish

Little’s Law: Inventory = flow rate * flow time

Capacity = 1/processing time Flow rate = minimum {demand, process capacity}

Utilization = flow rate/capacity Cycle time = 1/flow rate Time to make Q units = cycle time * Q

Time through empty worker paced system = sum of all processing times. Time through an empty machine paced
system = # of stations * cycle time Time to finish X units starting with an empty system = time through an empty
process + [(X-1)* cycle time]

Bottleneck has lowest capacity (?)

Demand Matrix: for each resource, i, and for each flow unit of type j: D(i,j) = number of flow units of type j that
need to be processed at resource I (Each row in a demand matrix corresponds to a resource and each column
corresponds to a flow unit type. Input 0s when customer type j doesn’t need attention from resource i)

Total demand rate for resource i = ∑jD(i,j)

Implied utilization = total demand at the resource / capacity at the resource (Can exceed 100% - any resource
that goes over 100% means it cannot meet the demand)

How to find the bottleneck and flow rates in a process with multiple units

1. demand matrix 2.total demand rate for resource 3. Capacity level 4. Implied utilization (the resource with the
highest is the bottleneck) 5. Flow rate (if the implied utilization is less than or equal to 100%, the deman rates are
the flow rates, if they are higher, divide the demand rate by the highest implies utilization)

Attrition losses: all flow units start at the same resource but then drop out of the process at different points

number of units to get Q good units = Q/process yield

number of units to get Q good units out of the resource = Q/yield at resource

capacity in good units per unit of time = 1 good unit per unit of time / implied utilization

Yield = flow rate of good output at the resource / flow rate of input = 1 – (flow rate of defects at the resource) /
flow rate of input

Process yield = flow rate of good output at the process / flow rate of input to the process = 1 – (flow rate of
defects at the process) / flow rate of input to the process
Process yield = y1*y2*y3*……………..ym (m is the number of resources in the sequence and y i is the yield of the ith
resource)

How to find the bottleneck and capacity with attrition loss

1. demand matrix 2. Capacity level for each resource 3. Implied utilization (highest = bottleneck) 4. Capacity in
good units = 1 good unit per unit of time/implied utilization

capacity expressed in available time per hour at resource i = number of workers (i) *60 [minutes / hours]

Workload matrix = WJ (i,j) = number of flow units of type j that need to be processed by resource i *processing
time that resource i takes for customer type j

Demand rate for resource j in minutes of work =∑jD(i,j)

Implied utilization = demand rate in minutes of work / capacity in minuts of work

How to find the bottleneck and flow rates with workflow dependence

1. workload matrix 2. Add workload for each 3. Available time at each resource 4. Implied utilization (above)
(highest = bottleneck) 4. Capcity= 1 unit per unit of time/implied utilization

Queue growth rate = demand – capacity Length of queue at time T = T * queue growth rate

Time to serve Qth person in the group = Q/capacity

Time to serve person arriving at time T = T*((Demand/capacity)- 1)

Average time to serve a unit = (1/2)*T*((demand/capacity)-1)

Peak-load pricing (adjust demand): charging more during your busiest times

Off peak discount (adjust demand): offering a discount during the nonbusy period in order to attract customers
for those times instead of the peak times

Pre processing strategy (adjust capacity): reducing the amount of work needed to process a customer during the
peak period by moving it to do for the off peak period

Arrival process: the flow of customers arriving to the system Interarrival time: the time between customer
arrivals to the system (a = average interarrival time)

Service process: the flow of customers when they are being served Processing time: the time a customer spends
with a server (p = average processing time)

Coefficient of variation: the ratio of standard devation to the average

CVa= standard devation of arrival process/a CVp= standard devation of service process/p

For single server - Utilization = p/a - Average time a customer waits in the queue (Tq) = p * (utilization/(1-
utilization)) * ((CVa2 + CVp2)/2) -Predicting the number of customers waiting and in service Iq = Tq/a Ip = p/a

For multiple servers (m = number of servers) - Utilization= p/(a*m) - Predicting waiting time Tq = (p/m) *
(utilization√2(m+1)-1/(1-utilization)) * ((CVa2 + CVp2)/2) - Predicting number of customers waiting and in service I q =
Tq/a Ip = p/a

Unstable queue: demand rate > capacity Stable queue: demand rate < capacity
Pooled queue system: all demand is shared amongst all the servers Separate queue system: demand is
immediately divided amongst servers and the customer is served only by their designated server