“An analytical model to increase air volumes and minimize ... · RESEARCH PROJECT PRESENTATION...
Transcript of “An analytical model to increase air volumes and minimize ... · RESEARCH PROJECT PRESENTATION...
NIKLAS BLOMBERG & RAMON GRAS MIT Master of Engineering in Logistics MIT Research Fest Final Presentation MAY 21 2015
“An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation”
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
AGENDA
1.- PROBLEM DEFINITION
2.- BUILDING THE DATABASE: KEY VARIABLES
3.- INTEGRATED DATA MANAGEMENT SYSTEM
4.- DATA SCARCITY AND SIMULATION
5.- METRICS DESIGN
5.1.- PHYSICAL PROPERTIES
5.2.- PROFITABILITY
6.- VISUALIZATION TOOLS
7.- CONCLUSIONS & FURTHER RESEARCH
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
1.- PROBLEM DEFINITION
Airfreight forwarding companies must develop accurate simulation tools to assess the attractiveness of each bidding process, to decide whether to participate in air cargo tenders of buying and reselling of air cargo space and how to define the optimal commercial strategy. “1:6” weight/volume ratio: establishes that whenever the cargo ratio is different from 1:6 (1m3:167 kg) forwarders must pay the highest rate: either volume or weight.
Chargeable Weight (carrier/airline) α kg
Rate paid to carrier ρ $/kg
Chargeable Weight (shipper) σ kg
Rate charged to shipper β $/kg
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
To increase competitiveness, 3PL companies and air freight forwarder firms can improve their consolidation techniques, to combine in the same load cargo with compatible densities. The availability of robust analytical resources will allow airfreight industry companies to increase their rate of financial success, in terms of enhancing both efficiency (by increasing air volumes) and profitability (by minimizing the Net Achieved Rate).
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
1.- PROBLEM DEFINITION
Main operational constraints for air cargo:
The most profitable business opportunities consist of combining in the same load different products with compatible densities through consolidation. Main target: to come close as possible to the desired 1:6 ratio for the DIM Factor, to minimize the average price per load.
Volumetric capacity Vt M3
Density δ Kg/M3
Net Achieved Rate = NAR = α𝑖·𝑛
𝑘=1ϱ𝑖
σ𝑖𝑛
𝑘=1
Profits Π = σ𝑖 ·
𝑛
𝑘=1
β𝑖 − α𝑖 ·
𝑛
𝑘=1
ϱ𝑖
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
1.- PROBLEM DEFINITION
Main questions this thesis addresses: 1.- Which are the most attractive incoming bids for any given o-d lane? 2.- To what extend is a particular current bid under consideration compatible with our current portfolio / current business for a given o-d lane to maximize the air volume and density usage and profitability surplus derived from consolidation? 3.- How to choose the rates so that we maximize profitability while still being competitive when addressing RFQs? 4.- How to depict such questions through visualization tools in an attractive and intuitive way to be able to rapidly address decision-making when facing air cargo transportation tenders?
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
THESIS OUTLINE
This thesis develops an analytical model based on meaningful metrics to provide airfreight forwarders with an accurate and solid forecasting tool to: 1.- select the most attractive bids under consideration, those that best match with their current portfolio/current business in terms of air volume usage and density efficiency for a given origin-destination lane. 2.- predicts breakeven rates to guide decision-making when addressing tenders to increase profitability by minimizing the Net Achieved Rate. 3.- visualization tools to help air freight forwarding companies to improve their understanding and depiction of their current situation in order to design their commercial strategy.
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
BOUNDARIES OF THE RESEARCH PROJECT
To narrow the scope of the project we applied the methodology developed exclusively to the • 9 customers from different industries, and • 9 target lanes we are focusing on: the routes between three Chinese cities
(origin) and three American cities (destination).
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
Origin (China) Destination (USA)
City Airport Code City Airport Code
Beijing PEK New York JFK
Hong Kong HKG Chicago ORD
Shanghai PVG Los Angeles LAX
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
BOUNDARIES OF THE RESEARCH PROJECT
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
Air Rates Customer ID Industry
C1, C9 Lifestyle/clothing
C2, C6 Retail - clothing
C3, C4 Industrial
C5, C7 Technology - hardware
C8 Toy industry
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
LITERATURE REVIEW
Main research gap: The current state-of-the-art research covers different areas with respect to the operational aspects of the consolidation problem, but does not present a methodology to assess through visualization tools the attractiveness of current bids / bids under consideration with respect to the commercial strategy.
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
Huang, Kuancheng; Chi, Wenhou. A Lagrangian relaxation based heuristic for the consolidation problem of airfreight forwarders. Transportation Research Part C 15 (2007) 235–245. Li, Zichao; Bookbinder, James H.; Elhedhli, Samir. Optimal shipment decisions for an airfreight forwarder: Formulation and solution methods. Transportation Research Part C 21 (2012) 17–30. Wong, Wai Hung; Leung; Lawrence C.; Hui, Yer Van. Airfreight forwarder shipment planning: A mixed 0–1 model and managerial issues in the integration and consolidation of shipments. European Journal of Operational Research 193 (2009), p86–97. Han, Dong Ling; Tang, Loon Ching; Huang, Huei Chuen. A Markov model for single-leg air cargo revenue management under a bid-price policy. European Journal of Operational Research. 200 (2010) p 800–811.
Surfaces of expected revenue with respect to bid prices.
Source: Han, Tang & Huang (2010)
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
2.- BUILDING THE DATABASE: KEY VARIABLES
Main given variables for consolidation of air cargo:
Origin = Origin Gateway for a certain bid
Destination = Destination Gateway for a certain bid
o-d lane = origin-destination pair, for each lane
Shipper = the company who ships cargo to the consignee. The shipper in the
context of this thesis is the client of the air freight forwarder
Carrier = the airline, the legal entity that is in the business of transporting goods
for hire
Request For Quotation: the tender, a negotiating approach whereby the buyer asks for a price quotation from a potential supplier from specific transportation services that a buyer needs over a certain time and at a fixed price
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
2.- BUILDING THE DATABASE: KEY VARIABLES
Main given variables for consolidation of air cargo: γ𝑤 = gross weight, the total weight in kg for a certain bid υ = volume, the total volume for a certain bid in CBM (cubic meters, 𝑚3) υ𝑤 = volumetric weight, the total volume for a certain bid, normalized using the 1:6 WV Ratio
δ =γ𝑤
υ density of cargo, as gross weight per volumetric unit, in
kg
𝑚3
τ𝑟 = Target Rate for a given o-d lane, reference provided by the shipper
DIM Factor = 1:X Weight/Volume Ratio for a certain bid, being X = 1000γ𝑤υ
=1000·𝜈
γ𝑤
ω𝜈 = 1:6 Weight/Volume Ratio for air freight, equivalent to the maximum density
(upper bound, threshold) permitted in a certain shipment on average: 167 kg
𝑚3
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
3.- INTEGRATED DATA MANAGEMENT SYSTEM
To build a complete, accurate and well-structured database,
by creating and integrated, holistic database
that combines datasets created by different teams
can be a challenge for maneuverability of records and data integrity.
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
3.- INTEGRATED DATA MANAGEMENT SYSTEM
A database management system such as SQL Server permits to integrate the different datasets from different professional teams
to help air freight forwarders to acquire a new level of proficiency in data management,
by implementing a set of queries to merge pieces of information from different datasets and teams, manipulating data in an efficient way,
and creating significant metrics to guide decision-making.
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
Datasets
DMS with embedded
analytical model
Visualization tools
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
3.- INTEGRATED DATA MANAGEMENT SYSTEM
SELECT OriDes, Origin, Destination, COUNT(id) AS [#], SUM(revenue) AS [revenue], SUM(cost) AS [cost], SUM(revenue)-SUM(cost) AS [profit], (SUM(revenue)-SUM(cost))/SUM(revenue) AS [margin], SUM(kg_per_year) AS [shipperchkg], SUM(carrierchkg) AS [carrierchkg], SUM([gross weight]) AS [totgross], SUM([Vol weight]) AS [totvol], 1-dbo.InlineMax(SUM([gross weight]), SUM([Vol weight]))/SUM([kg_per_year]) AS [cons], SUM(cost)/SUM(kg_per_year) AS [NAR], (1000/(167*(SUM([gross weight])/SUM([vol weight])))) AS [1:], (SUM(revenue)-SUM(cost))/SUM(m3) AS [profit/m3], SUM(kg_per_year * [total cost])/SUM(kg_per_year) AS [WeightedAvgRate], SUM([vol weight]) AS [VW], SUM([gross weight]) AS [GW], (CASE WHEN SUM([vol weight])-SUM([gross weight])>0 THEN SUM([vol weight])-SUM([gross weight]) WHEN SUM([vol weight])-SUM([gross weight])<=0 THEN 0 END) AS [freeGW], (CASE WHEN SUM([gross weight])-SUM([vol weight])>0 THEN SUM([gross weight])-SUM([vol weight]) WHEN SUM([gross weight])-SUM([vol weight])<=0 THEN 0 END) AS [freeVW], SUM([gross weight])/(dbo.InlineMax(SUM([gross weight]), SUM([Vol weight]))) AS [grosseff], SUM([vol weight])/(dbo.InlineMax(SUM([gross weight]), SUM([Vol weight]))) AS [voleff]
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
3.- INTEGRATED DATA MANAGEMENT SYSTEM
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
CUSTOMERS&ROUTES PHYSICAL PROPERTIES
TENDERS & RATES
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
4.- DATA SCARCITY AND SIMULATION
Data scarcity with respect to gross weight and density / DIM Factor
meant a significant flaw that freight forwarders may have as a result of not having a robust and integrated data management system.
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
0
5000
10000
15000
20000
25000
30000
Total rows Total cost > 0 Kg > 0 Max target > 0 Target and kg > 0
Available data (inputs/rows)
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
4.- DATA SCARCITY AND SIMULATION
Data scarcity with respect to gross weight and density / DIM Factor
meant a significant flaw that freight forwarders may have as a result of not having a robust and integrated data management system.
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
0
200
400
600
800
1000
1200
1400
1600
9 routes 9 Total cost > 0 9 Kg > 0 9 Max target > 0 9 Target and kg > 0
Available data - 9 customers & routes
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
4.- DATA SCARCITY AND SIMULATION
Data scarcity with respect to gross weight and density / DIM Factor
meant a significant flaw that freight forwarders may have as a result of not having a robust and integrated data management system.
Thus, we opted to simulate the missing values performing statistical simulations
by using the statistical distribution which presents a better adjustment of the sample values for both variables.
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
0
50
100
150
200
250
300
350
1 2 3 4 5 6 7 8 9
10
11
12
13
14
15
16
17
18
19
20
DIM Factor
Comparison between RFI03 and Expected Lognormal - Frequency Distribution
Frequency
Expected Frequency
0.00%10.00%20.00%30.00%40.00%50.00%60.00%70.00%80.00%90.00%
100.00%
1
2.5 4
5.5 7
8.5 10
11.5 13
14.5 16
17.5 19
Cu
mu
lati
ve F
req
uen
cy
DIM Factor
RFI03 Cumulative Distribution
CumulativeFrequency
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
DIM Factor adjustment Gross Weight adjustment
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
4.- DATA SCARCITY AND SIMULATION
The results of tests of goodness of fit according to the Kolmogorov-Smirnov
methodology showed that the DIM Factor was adjusted using a Lognormal distribution
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
0
100
200
300
400
500
600
1 2 3 4 5 6 7 8 9 101112131415161718192021222324252627282930
Fre
qu
en
cy
DIM Factor
Lognormal distribution Goodness of Fit RFI01 & RFI03b Merged
Frequency
Expected Frequency
Lognormal Distribution
𝑓(𝑥; 𝜇, 𝜎) =1
𝑥𝜎 2𝛱𝑒−(ln 𝑥 −𝜇)2
2𝜎2
with parameters
μ=2.2633
σ=0.33116
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
4.- DATA SCARCITY AND SIMULATION
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
0
200
400
600
800
1000
1200
1400
1600
1800
2000
100
600
1100
1600
2100
2600
3100
3600
4100
4600
5100
5600
6100
6600
7100
7600
8100
8600
9100
9600
Fre
quen
cy
Gross Weight
Frequency Distribution Goodness of Fit - Gamma 3P
Frequency
Expected Frequency
whereas the gross weight γ𝑤 values show the highest levels of goodness of fit with the Generalized Gamma Distribution.
Generalized Gamma Distribution
𝑓(𝑥; 𝑎, 𝑑, 𝑝) =
𝑝
𝑎𝑑·𝑥𝑑−1
Γ(𝑑/𝑝)𝑒−(𝑥
𝑎)𝑝
with parameters
a=0.38684 b=1906.8 g=100.0
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
5.- METRICS DESIGN
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
5.1.- PHYSICAL PROPERTIES
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
Average density of shipments and aggregate business opportunities by origin
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
5.2.- PROFITABILITY
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
Profitability Performance Metrics
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
5.2.- PROFITABILITY
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
Increased consolidation potential permits to boost overall profitability, as a result of being able to charge twice a rate to different shippers when combining heavier and lighter cargo in the same load. We focused primarily on the surplus that consolidation permits to achieve by leveraging the increased Level of Aggressiveness on Bidding (Г) when addressing the last rounds of the RFQ.
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
5.2.- PROFITABILITY
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
To remain profitable for the whole current business/portfolio for a given route, air freight forwarders can use as reference values: Break-even Rate for Constant Profits (βπ), rate guarantees the same profits in absolute numbers overall per lane. It can be used as a lower bound or minimum value to determine the rate (β) to submit to the shipper; Constant Profitability per Lane Rate (βλ), guarantees the same level of profitability as the current business.
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
ANALYTICAL MODEL EMBEDDED IN SQL SERVER
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
CUSTOMERS&ROUTES PHYSICAL PROPERTIES
TENDERS & RATES
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
6.- VISUALIZATION TOOLS
Visualization tool 01: Consolidation Delta with respect to Portfolio DIM Factor (Hong Kong – LA route)
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
6.- VISUALIZATION TOOLS
Visualization tool 02: Consolidation Delta Δ with respect to Shipper Ch kg σ
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
6.- VISUALIZATION TOOLS
By selecting one bid in particular, the visualization tool displays the whole sets of meaningful variables and metrics
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
6.- VISUALIZATION TOOLS
Display of underlying data table in Tableau
An analytical model to increase air volumes and minimize the Net Achieved Rate in air freight transportation
RESEARCH PROJECT PRESENTATION NIKLAS BLOMBERG | RAMON GRAS
7.- CONCLUSIONS
KEY INSIGHTS 1.- Building a complete and well-structured database using a scalable and repeatable process permits to gather, manage and interpret relevant data from past tenders and allows air freight forwarders for having an accurate depiction of levels of air usage efficiency and profitability. 2.- Creating meaningful metrics is crucial to assess the attractiveness of incoming bids. 3.- Designing intuitive visualization tools helps to refine decision-making when addressing air cargo tenders. FURTHER RESEARCH To align the commercial strategy with operational procedures built upon constraints and limitations that occur at the consolidation execution level. E.g., to combine the analytical model embedded in SQL Server with Lagrangian-relaxation heuristics to solve through software such as CPLEX the optimization of the consolidation problem at an operational level