BMGT43290 Building Sustainable Supply Chain Advantage

Littlefield Simulation Report

Group: Team4

Kush Anant Jain, 23205522

Aabhash Pansari, 23206252
Alicia Breslin, 20428564
Viola Eichenseer, 23202651
Ria , 23200787

We declare that all materials included in this essay/report/project/dissertation is the result of

our own work and that due acknowledgement has been given in the bibliography and
references to ALL sources be they printed, electronic or personal.

General strategy

To succeed in the simulation, we developed a strategy focused on analysing inventory levels and
consumer feedback. Using data from the previous 50 days, we aimed to maintain lead times under
one day, prevent machine utilization from reaching 100% to avoid bottlenecks, and ensure revenue
didn't drop below $1000 for three consecutive days. This approach required a deep understanding of
our production capabilities and the agility to adapt to changes. Initially, our factory's machine
distribution across tuning, testing, and board stuffing stations was designed to ensure smooth
operations and prevent any station from becoming a bottleneck. However, our strategies evolved as
we gained more operational insights during the simulation.

WEEK 1

Capacity Decisions and Rationale

1. Initial Setup:
 Station 1 (Board Stuffing): The machine count was set to 3.
 Station 2 (Testing): The machine count was set to 2.
 Station 3 (Tuning): The machine count was set to 2.
Kits Queued at Each Station
a. Station 1 (Board Stuffing)
 Plot Analysis: faced initial bottlenecks, suggesting initial bottlenecks.
 Interpretation: Early peaks indicate insufficient initial capacity or inefficiencies in
the board stuffing process, possibly due to a learning curve or initial misalignment
between kit supply rates and processing capacity.
b. Station 2 (Testing)
 Plot Analysis: showed fewer, lower peaks, suggesting sporadic overloads.
 Interpretation: The lesser degree of queuing compared to Station 1 might imply a
slightly better alignment with incoming and outgoing kit flow, yet the station still
faced periods of overload, influencing overall cycle times.
c. Station 3 (Tuning)
 Plot Analysis: experienced significant bottlenecks, affecting overall lead times.
 Interpretation: This station's capacity issues could have delayed downstream
processes and affected overall lead times significantly.

Further Analysis of Capacity Management Decisions

Initial Scheduling Rule (Priority to Step 4 - pri4)

 Implementation Time: Day 74

 Decision Rationale: Prioritizing Step 4 was likely aimed to expedite the processing
of units at their final testing phase, ensuring quicker completion of the final product
and potentially reducing lead times.

 Impact Assessment: This decision would have been particularly effective if Step 4
was a bottleneck or if fulfilling orders quickly was crucial under the simulation
conditions.

Switch to First-In, First-Out (FIFO)

 Implementation Time: Day 95

 Decision Rationale: Switched to FIFO to simplify operations after finding the priority
rule ineffective and overly complex.
 Impact Assessment: FIFO made scheduling fairer and reduced wait times and queues,
especially during peak periods.

Recommendations Based on Data Analysis

 Station 1 and 3: Increase machine counts during peak periods, using dynamic scheduling for
optimal resource allocation.
 Station 2: Adjust capacity as needed by closely monitoring the input from Station 1 to
maintain throughput and minimize queuing.

Forecasting
 Number of jobs accpeted each day
10

6
f(x) = 0.0511278195488722 x + 2.10676691729323
5

0
50 55 60 65 70 75

Analysis of the Forecasting Method

1. Trend Line Equation: The trend line is represented by the equation
𝑦=0.0641𝑥+0.995y=0.0641x+0.995. The slope (0.0641) indicates a gradual increase in job
acceptance.
2. Forecasting Future Demand: Using this trend line, you can forecast expected job
acceptance for days 51 to 70. This will help in anticipating future demand and adjusting your
operational strategies accordingly.

3.
Application of Forecast Data
1. Inventory Management: The increase in job acceptance indicates a need to adjust reorder
points and order quantities to meet rising demand and prevent stockouts while maintaining
efficiency.
2. Capacity Planning: As job acceptance grows, evaluate if the current capacity can meet the
increased demand without causing delays or queues.

Mistakes
In the simulation's later stages, we removed a machine from each station on days 166.08, 154.98, and
144.48 to optimize costs based on current demand and efficiencies. These reductions were made too
soon leading to decreased revenue. Reducing machine numbers aimed to cut costs but resulted in
longer lead times due to insufficient capacity for variable demand, increasing the risk of penalties for
delayed orders. It limited our flexibility, hindering our response to sudden demand surges and
ultimately impacting our revenue.

WEEK 2

Contract selection

There are three contracts available in the Littlefield Technologies (LT) simulation: Contracts 1, 2, and
3. Each contract has a different pricing and lead time. Despite its higher cost, Contract 3 turns out to
be the best choice after analysis. Because of its reduced lead time, it can respond to consumer
demands more quickly, which may boost customer satisfaction and maybe improve revenue. In
addition, a shorter lead time reduces the possibility of overproduction or inventory accumulation,
which ultimately saves money. In summary, Contract 3 is the most favourable option for Littlefield
Technologies (LT) in the simulation since it offers the best lead time and benefit combination.

At the start of Day 51, we decided to go to Contract 3 without properly accounting for the variable
lead time, which proved to be a mistake. Our revenue thus decreased, and during the course of the
following simulated days, we found ourselves going back to contract 2. We saw a decrease in revenue
from Day 61 to Day 146 as a result of this.

We were able to shorten the lead time by making inventory modifications and buying extra machines
at various stations. In the end, we decided to stick with Contract 3 for the duration of the simulation.

Capacity Decisions

When simulation 2 started, we saw that station 2's use peaked for five days in a row. As a result, we
purchased one machine on day fifty, and our usage never decreased after that. This was on day fifty-
four. After purchasing the machine, station 2's average usage was 0.55 through day 268. Similarly on
day 50, we purchased one machine for station 1 after noting that, five days before to the purchase, its
utilization had reached 1.0. This increased the number of machines at station 1 to four in total. After
acquisition, the average utilization was 0.55 until the completion of the simulation. It eventually
decreased to 1.0, even though it had increased three times meaning further machines didn’t need to be
acquired.

We added a machine to station 3 after noting its full utilization on day 50, bringing its total to two
machines and lowering the average utilization to 0.52. Post-purchase, stations 1 and 2 also had two
machines each, while station 3 had three. In the second simulation, our timing in purchasing the
correct number of machines for each station was poorly judged, impacting our evaluation.

Capacity needs are driven by average demand, which can occasionally lead to queues if demand
spikes above the average, or dissipate if it falls below. During the first fifty days, our average
demand was 12.24 lots per day. Industry standards suggest a utilization rate of 70% to 85%. We
targeted a utilization rate of 85% for our stations, meaning the 12.24 lots per day represented 85% of
our station's total capacity.

0.85 = 12.24 units/days Station capacity/day = 12.24 Station capacity/day = 14.4 units/day
Station Capacity/day 0.85

Therefore, required station capacity = 14.4 lots / day

Hence, No. of machines required will be:
Station1:14.4/3.80=3.78 =4machines(upper limit)
Station2:14.4/11.41=1.26 =2machines(upper limit)
Station3:14.4/11.41=1.26 =2machines(upper limit)

Machine Priority at Station 2: We observed that there is always no backlog. Therefore, shifting the
priority is not necessary. The money made will remain unchanged as a result of this choice.

Inventory Decision

Despite the fact that keeping inventory has no physical cost, there are financial costs to consider. With
a fixed cost per order of $1000 and an annual demand rate of 4467 units/year (12.24 lots/day), the
order point is calculated using the formula below. $60 represents the yearly holding cost, or the
interest lost as a result of the decrease in the cash balance (600 divided by 0.1, or 10%). 386 lots,
rounded to 385.90, is the EOQ value. The reorder point is 61 (rounded to 60.8) lots (3660 kits),
assuming a 4-day delivery time and a standard deviation of 3.6 computed from the demand rate.

We recommend setting the reorder point to zero and modifying the order point to 36,720 lots—the
average inventory needed for 50 days—for the final 50 days of the simulation. By making this choice,
we avoid wasting money by ensuring that we have little inventory on the last day.

Calculation for inventory

1. Order quantity
Demand Rate=12.24 lots/day
Fixed cost per order=$1000
Annual Interest rate on cash balance=10%

Formula for EOQ =

=
=385.90units

(1unit=60 kits)=385.90*60
=23,154 kits

2. Reordering point
Reorder quantity=(average lead time x average demand)+safety stock
= (4 daysx12.24 orders per day)+safety stock

Safety stock=NORM.S.INV(service level) x𝝈√𝐿

Service Level=95%
Standard deviation of demand=(𝝈)=STDDEV(Demand rate)=3.38
Safety Stock =11.11 lots
Reorder point=(4x12.24)+11.11=60.7 lots=61 approximately=3660 lots

3. Order and reorder point for last 50 days

Approx. inventory used per day =12.24*60 =734.4 kits
Inventory used in 50 days =50*12.24=612 lots=36,720 kits.

1. Graphs

1.Queue for Station 1 2.Station 1 Utilization

3.Queue for Station 2 4.Station 2 Utilization

5.Queue for Station 3 6.Station 3 Utilization

 Mistakes recognised in Week 2

The initial choice of Contract 3 without proper lead time management led to lower than expected
revenue. We learned the importance of closely monitoring simulation parameters and making timely
adjustments to maximize outcomes