390.Can SAS Predict Airline Disasters Before They Happen?

 Can SAS Predict Airline Disasters Before They Happen?

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

HERE IN THIS PROJECT WE USED THESE SAS STATEMENTS —
DATA STEP | PROC SQL |  PROC PRINT | PROC SGPLOT | MACROS | PROC CORR | PROC MEANS | PROC FREQ | PROC UNIVARIATE | APPEND | PROC SORT | MERGE | PROC DATASETS DELETE | DATA FUNCTIONS

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

INTRODUCTION

The airline industry operates in one of the most time-sensitive and safety-critical environments in the world. Every day, thousands of aircraft take off and land across global airspace, and behind each flight is a complex system of human scheduling, regulatory compliance, and fatigue management.

Among all operational components, airline crew scheduling is one of the most crucial and sensitive functions. Pilots, co-pilots, and cabin crew must follow strict regulations regarding:

·       Maximum allowable flight hours

·       Minimum mandatory rest hours

·       Fatigue and stress thresholds

·       Aircraft type certifications

·       Regulatory compliance with aviation authorities

Any failure in this system can result in:

·       Flight delays and cancellations

·       Crew burnout and attrition

·       Regulatory penalties and audits

·       Severe safety risks and reputational damage

Therefore, modern airlines rely heavily on data analytics systems to monitor and optimize crew schedules in real time.

This project simulates a real-world airline crew analytics platform built using SAS, where we design and analyze a crew scheduling dataset using:

·       PROC SQL

·       PROC MEANS

·       PROC UNIVARIATE

·       PROC FREQ

·       PROC CORR

·       PROC SGPLOT

·       SAS Macros

·       Date functions (MDY, INTCK, INTNX)

·       Data integration (SET, MERGE, APPEND, TRANSPOSE)

·       Data quality functions (character and numeric functions)

The system not only performs statistical analysis but also implements business logic for fatigue detection, utilization classification, and fraud/risk flagging, exactly as used in real airline operations.

 

BUSINESS CONTEXT

In real airline companies such as Air India, Emirates, Indigo, Qatar Airways, Lufthansa, and British Airways, crew management departments continuously answer critical questions like:

·       Which crews are overworked?

·       Are minimum rest hours being violated?

·       Which aircraft types generate more fatigue?

·       Is any crew member manipulating duty logs?

·       How does flight load impact human performance?

These questions cannot be answered manually. They require:

·       Automated data pipelines

·       Statistical monitoring

·       Visualization dashboards

·       Compliance alerts

·       Risk analytics

This SAS project acts as a mini airline operations control center where management can:

·       Monitor fatigue levels

·       Detect unsafe schedules

·       Compare utilization patterns

·       Identify high-risk crews

·       Ensure regulatory compliance

 

TABLE OF CONTENTS

1.     Introduction

2.     Business Context

3.     Dataset Design

4.     Variable Definitions

5.     Data Creation in SAS

6.     Date Handling and Formats

7.     Utilization Classification Macro

8.     Fraud and Fatigue Detection Macro

9.     PROC SQL – Summary Analytics

10.  PROC MEANS – Descriptive Statistics

11.  PROC UNIVARIATE – Distribution Analysis

12.  PROC FREQ – Categorical Insights

13.  PROC CORR – Relationship Analysis

14.  PROC SGPLOT – Visualization

15.  Data Integration (SET, MERGE, APPEND)

16.  Data Reshaping (TRANSPOSE)

17.  Character Functions Usage

18.  Numeric Functions Usage

19.  Data Cleaning and Deletion

20.  Conclusion

Why this project is realistic

This project exactly mirrors how airline analytics systems are built in real companies like:

·       Emirates

·       Indigo

·       Air India

·       Qatar Airways

Crew management teams track:

·       Working hours

·       Rest time

·       Fatigue levels

·       Regulatory compliance

·       Fraud or data manipulation

Dataset design logic

Each variable has operational meaning:

Variable	Meaning
Crew_ID	Unique employee ID
Name	Crew member name
Aircraft_Type	Aircraft certification
Duty_Date	Date of duty
Flight_Hours	Total flying hours
Rest_Hours	Total rest hours
Delay_Incidents	Delays caused
Fatigue_Score	Human performance metric
Utilization_Pct	Productivity ratio
Utilization_Status	High/Medium/Low
Fraud_Flag	Risk indicator

 

STEP 1 – CREATE MAIN DATASET

data crew_schedule;

    input Crew_ID $ Name:$13. Aircraft_Type $ Duty_Date :date9.  

          Flight_Hours Rest_Hours Delay_Incidents Fatigue_Score;

          Utilization_Pct = Flight_Hours / 12;

    format Duty_Date date9. Utilization_Pct percent8.2;

    datalines;

C001 Ramesh_A320 A320 01JAN2026 8 10 1 60

C002 Sita_B737 B737 02JAN2026 10 8 2 75

C003 Arjun_A320 A320 03JAN2026 6 12 0 45

C004 Meera_A350 A350 04JAN2026 11 7 3 85

C005 Kiran_B787 B787 05JAN2026 9 9 1 65

C006 Pooja_A320 A320 06JAN2026 12 6 4 90

C007 Rahul_B737 B737 07JAN2026 7 11 0 50

C008 Neha_A350 A350 08JAN2026 10 8 2 70

C009 Vinay_B787 B787 09JAN2026 5 13 0 40

C010 Kavya_A320 A320 10JAN2026 9 9 1 68

C011 Ajay_B737 B737 11JAN2026 11 7 3 82

C012 Divya_A350 A350 12JAN2026 8 10 1 60

C013 Sanjay_B787 B787 13JAN2026 12 6 4 92

C014 Anu_A320 A320 14JAN2026 6 12 0 48

C015 Rohit_B737 B737 15JAN2026 10 8 2 75

C016 Sneha_A350 A350 16JAN2026 7 11 0 55

;

run;

proc print data=crew_schedule;

run;

OUTPUT:

Obs	Crew_ID	Name	Aircraft_Type	Duty_Date	Flight_Hours	Rest_Hours	Delay_Incidents	Fatigue_Score	Utilization_Pct
1	C001	Ramesh_A320	A320	01JAN2026	8	10	1	60	66.67%
2	C002	Sita_B737	B737	02JAN2026	10	8	2	75	83.33%
3	C003	Arjun_A320	A320	03JAN2026	6	12	0	45	50.00%
4	C004	Meera_A350	A350	04JAN2026	11	7	3	85	91.67%
5	C005	Kiran_B787	B787	05JAN2026	9	9	1	65	75.00%
6	C006	Pooja_A320	A320	06JAN2026	12	6	4	90	100.0%
7	C007	Rahul_B737	B737	07JAN2026	7	11	0	50	58.33%
8	C008	Neha_A350	A350	08JAN2026	10	8	2	70	83.33%
9	C009	Vinay_B787	B787	09JAN2026	5	13	0	40	41.67%
10	C010	Kavya_A320	A320	10JAN2026	9	9	1	68	75.00%
11	C011	Ajay_B737	B737	11JAN2026	11	7	3	82	91.67%
12	C012	Divya_A350	A350	12JAN2026	8	10	1	60	66.67%
13	C013	Sanjay_B787	B787	13JAN2026	12	6	4	92	100.0%
14	C014	Anu_A320	A320	14JAN2026	6	12	0	48	50.00%
15	C015	Rohit_B737	B737	15JAN2026	10	8	2	75	83.33%
16	C016	Sneha_A350	A350	16JAN2026	7	11	0	55	58.33%

STEP 2 – UTILIZATION CLASSIFICATION MACRO

%macro utilization_class;

data crew_schedule;

    set crew_schedule;

    length Utilization_Status $8.;

    if Flight_Hours >= 10 then Utilization_Status = "HIGH";

    else if Flight_Hours >= 7 then Utilization_Status = "MEDIUM";

    else Utilization_Status = "LOW";

run;

proc print data=crew_schedule;

 var Crew_ID Name Aircraft_Type Flight_Hours Utilization_Status;

run;

%mend;

%utilization_class;

OUTPUT:

Obs	Crew_ID	Name	Aircraft_Type	Flight_Hours	Utilization_Status
1	C001	Ramesh_A320	A320	8	MEDIUM
2	C002	Sita_B737	B737	10	HIGH
3	C003	Arjun_A320	A320	6	LOW
4	C004	Meera_A350	A350	11	HIGH
5	C005	Kiran_B787	B787	9	MEDIUM
6	C006	Pooja_A320	A320	12	HIGH
7	C007	Rahul_B737	B737	7	MEDIUM
8	C008	Neha_A350	A350	10	HIGH
9	C009	Vinay_B787	B787	5	LOW
10	C010	Kavya_A320	A320	9	MEDIUM
11	C011	Ajay_B737	B737	11	HIGH
12	C012	Divya_A350	A350	8	MEDIUM
13	C013	Sanjay_B787	B787	12	HIGH
14	C014	Anu_A320	A320	6	LOW
15	C015	Rohit_B737	B737	10	HIGH
16	C016	Sneha_A350	A350	7	MEDIUM

Airlines classify crew into:

·       Overworked (HIGH)

·       Normal (MEDIUM)

·       Underused (LOW)

STEP 3 – FATIGUE / FRAUD DETECTION MACRO

%macro fraud_detection;

data crew_schedule;

    set crew_schedule;

    

    if Fatigue_Score > 80 and Rest_Hours < 8 then Fraud_Flag = "YES";

    else Fraud_Flag = "NO";

run;

proc print data=crew_schedule;

 var Crew_ID Name Aircraft_Type Fatigue_Score Rest_Hours Fraud_Flag;

run;

%mend;

%fraud_detection;

OUTPUT:

Obs	Crew_ID	Name	Aircraft_Type	Fatigue_Score	Rest_Hours	Fraud_Flag
1	C001	Ramesh_A320	A320	60	10	NO
2	C002	Sita_B737	B737	75	8	NO
3	C003	Arjun_A320	A320	45	12	NO
4	C004	Meera_A350	A350	85	7	YES
5	C005	Kiran_B787	B787	65	9	NO
6	C006	Pooja_A320	A320	90	6	YES
7	C007	Rahul_B737	B737	50	11	NO
8	C008	Neha_A350	A350	70	8	NO
9	C009	Vinay_B787	B787	40	13	NO
10	C010	Kavya_A320	A320	68	9	NO
11	C011	Ajay_B737	B737	82	7	YES
12	C012	Divya_A350	A350	60	10	NO
13	C013	Sanjay_B787	B787	92	6	YES
14	C014	Anu_A320	A320	48	12	NO
15	C015	Rohit_B737	B737	75	8	NO
16	C016	Sneha_A350	A350	55	11	NO

In real systems:

Some crew manipulate logs to:

·       Avoid night shifts

·       Claim overtime

·       Hide fatigue

Hence:

If fatigue is high + rest is low → flag

STEP 4 – PROC SQL ANALYSIS

proc sql;

    create table summary_sql as

    select Aircraft_Type,

           avg(Flight_Hours) as Avg_Flight,

           avg(Rest_Hours) as Avg_Rest,

           sum(Delay_Incidents) as Total_Delays

    from crew_schedule

    group by Aircraft_Type;

quit;

proc print data=summary_sql;

run;

OUTPUT:

Obs	Aircraft_Type	Avg_Flight	Avg_Rest	Total_Delays
1	A320	8.20000	9.80000	6
2	A350	9.00000	9.00000	6
3	B737	9.50000	8.50000	7
4	B787	8.66667	9.33333	5

SQL is used for:

·       Dashboards

·       KPI reports

·       Management decisions

STEP 5 – PROC MEANS

proc means data=crew_schedule mean min max std;

    var Flight_Hours Rest_Hours Fatigue_Score;

run;

OUTPUT:

The MEANS Procedure

Variable	Mean	Minimum	Maximum	Std Dev
Flight_Hours Rest_Hours Fatigue_Score	8.8125000 9.1875000 66.2500000	5.0000000 6.0000000 40.0000000	12.0000000 13.0000000 92.0000000	2.1975365 2.1975365 16.2090510

These are risk models:

·       High mean fatigue → unsafe operations

·       High variance → inconsistent scheduling

STEP 6 – PROC UNIVARIATE

proc univariate data=crew_schedule;

    var Fatigue_Score;

    histogram;

run;

OUTPUT:

The UNIVARIATE Procedure

Variable: Fatigue_Score

Moments
N	16	Sum Weights	16
Mean	66.25	Sum Observations	1060
Std Deviation	16.209051	Variance	262.733333
Skewness	0.04293774	Kurtosis	-1.0626313
Uncorrected SS	74166	Corrected SS	3941
Coeff Variation	24.466492	Std Error Mean	4.05226274

Basic Statistical Measures
Location		Variability
Mean	66.25000	Std Deviation	16.20905
Median	66.50000	Variance	262.73333
Mode	60.00000	Range	52.00000
		Interquartile Range	26.00000

Note: The mode displayed is the smallest of 2 modes with a count of 2.

Tests for Location: Mu0=0
Test	Statistic		p Value
Student's t	t	16.34889	Pr > |t|	<.0001
Sign	M	8	Pr >= |M|	<.0001
Signed Rank	S	68	Pr >= |S|	<.0001

Quantiles (Definition 5)
Level	Quantile
100% Max	92.0
99%	92.0
95%	92.0
90%	90.0
75% Q3	78.5
50% Median	66.5
25% Q1	52.5
10%	45.0
5%	40.0
1%	40.0
0% Min	40.0

Extreme Observations
Lowest		Highest
Value	Obs	Value	Obs
40	9	75	15
45	3	82	11
48	14	85	4
50	7	90	6
55	16	92	13

The UNIVARIATE Procedure

These are risk models:

·       High mean fatigue → unsafe operations

·       High variance → inconsistent scheduling

STEP 7 – PROC FREQ

proc freq data=crew_schedule;

    tables Utilization_Status Fraud_Flag Fatigue_Score;

run;

OUTPUT:

The FREQ Procedure

Utilization_Status	Frequency	Percent	Cumulative Frequency	Cumulative Percent
HIGH	7	43.75	7	43.75
LOW	3	18.75	10	62.50
MEDIUM	6	37.50	16	100.00

Fraud_Flag	Frequency	Percent	Cumulative Frequency	Cumulative Percent
NO	12	75.00	12	75.00
YES	4	25.00	16	100.00

Fatigue_Score	Frequency	Percent	Cumulative Frequency	Cumulative Percent
40	1	6.25	1	6.25
45	1	6.25	2	12.50
48	1	6.25	3	18.75
50	1	6.25	4	25.00
55	1	6.25	5	31.25
60	2	12.50	7	43.75
65	1	6.25	8	50.00
68	1	6.25	9	56.25
70	1	6.25	10	62.50
75	2	12.50	12	75.00
82	1	6.25	13	81.25
85	1	6.25	14	87.50
90	1	6.25	15	93.75
92	1	6.25	16	100.00

STEP 8 – PROC CORR

proc corr data=crew_schedule;

    var Flight_Hours Rest_Hours Fatigue_Score;

run;

OUTPUT:

The CORR Procedure

3 Variables:	Flight_Hours Rest_Hours Fatigue_Score

Simple Statistics
Variable	N	Mean	Std Dev	Sum	Minimum	Maximum
Flight_Hours	16	8.81250	2.19754	141.00000	5.00000	12.00000
Rest_Hours	16	9.18750	2.19754	147.00000	6.00000	13.00000
Fatigue_Score	16	66.25000	16.20905	1060	40.00000	92.00000

Pearson Correlation Coefficients, N = 16 Prob > |r| under H0: Rho=0
	Flight_Hours	Rest_Hours	Fatigue_Score
Flight_Hours	1.00000	-1.00000 <.0001	0.99148 <.0001
Rest_Hours	-1.00000 <.0001	1.00000	-0.99148 <.0001
Fatigue_Score	0.99148 <.0001	-0.99148 <.0001	1.00000

This tells management:

Does more flying cause fatigue?

If correlation is strong → reschedule crews.

STEP 9 – PROC SGPLOT

proc sgplot data=crew_schedule;

    scatter x=Flight_Hours y=Fatigue_Score;

    reg x=Flight_Hours y=Fatigue_Score;

run;

OUTPUT:

Executives do not read tables.

They want graphs:

Flight hours vs fatigue visually proves burnout.

STEP 10 – DATE FUNCTIONS

data crew_dates;

    set crew_schedule;

    

    Next_Duty = intnx('day', Duty_Date, 2);

    Gap_Days = intck('day', Duty_Date, Next_Duty);

run;

proc print data=crew_dates;

run;

OUTPUT:

Obs	Crew_ID	Name	Aircraft_Type	Duty_Date	Flight_Hours	Rest_Hours	Delay_Incidents	Fatigue_Score	Utilization_Pct	Utilization_Status	Fraud_Flag	Next_Duty	Gap_Days
1	C001	Ramesh_A320	A320	01JAN2026	8	10	1	60	66.67%	MEDIUM	NO	24109	2
2	C002	Sita_B737	B737	02JAN2026	10	8	2	75	83.33%	HIGH	NO	24110	2
3	C003	Arjun_A320	A320	03JAN2026	6	12	0	45	50.00%	LOW	NO	24111	2
4	C004	Meera_A350	A350	04JAN2026	11	7	3	85	91.67%	HIGH	YES	24112	2
5	C005	Kiran_B787	B787	05JAN2026	9	9	1	65	75.00%	MEDIUM	NO	24113	2
6	C006	Pooja_A320	A320	06JAN2026	12	6	4	90	100.0%	HIGH	YES	24114	2
7	C007	Rahul_B737	B737	07JAN2026	7	11	0	50	58.33%	MEDIUM	NO	24115	2
8	C008	Neha_A350	A350	08JAN2026	10	8	2	70	83.33%	HIGH	NO	24116	2
9	C009	Vinay_B787	B787	09JAN2026	5	13	0	40	41.67%	LOW	NO	24117	2
10	C010	Kavya_A320	A320	10JAN2026	9	9	1	68	75.00%	MEDIUM	NO	24118	2
11	C011	Ajay_B737	B737	11JAN2026	11	7	3	82	91.67%	HIGH	YES	24119	2
12	C012	Divya_A350	A350	12JAN2026	8	10	1	60	66.67%	MEDIUM	NO	24120	2
13	C013	Sanjay_B787	B787	13JAN2026	12	6	4	92	100.0%	HIGH	YES	24121	2
14	C014	Anu_A320	A320	14JAN2026	6	12	0	48	50.00%	LOW	NO	24122	2
15	C015	Rohit_B737	B737	15JAN2026	10	8	2	75	83.33%	HIGH	NO	24123	2
16	C016	Sneha_A350	A350	16JAN2026	7	11	0	55	58.33%	MEDIUM	NO	24124	2

Aviation runs on time. Without proper dates:

·       You cannot calculate rest periods

·       You cannot validate duty cycles

·       You cannot schedule shifts

Hence:

·       MDY() builds dates

·       INTCK() calculates gaps

·       INTNX() predicts future duties

STEP 11 – CHARACTER FUNCTIONS

data char_demo;

    set crew_schedule;

    

    Clean_Name = propcase(lowcase(strip(Name)));

    Full_Label = catx("-", upcase(Crew_ID), Aircraft_Type);

run;

proc print data=char_demo;

run;

OUTPUT:

Obs	Crew_ID	Name	Aircraft_Type	Duty_Date	Flight_Hours	Rest_Hours	Delay_Incidents	Fatigue_Score	Utilization_Pct	Utilization_Status	Fraud_Flag	Clean_Name	Full_Label
1	C001	Ramesh_A320	A320	01JAN2026	8	10	1	60	66.67%	MEDIUM	NO	Ramesh_a320	C001-A320
2	C002	Sita_B737	B737	02JAN2026	10	8	2	75	83.33%	HIGH	NO	Sita_b737	C002-B737
3	C003	Arjun_A320	A320	03JAN2026	6	12	0	45	50.00%	LOW	NO	Arjun_a320	C003-A320
4	C004	Meera_A350	A350	04JAN2026	11	7	3	85	91.67%	HIGH	YES	Meera_a350	C004-A350
5	C005	Kiran_B787	B787	05JAN2026	9	9	1	65	75.00%	MEDIUM	NO	Kiran_b787	C005-B787
6	C006	Pooja_A320	A320	06JAN2026	12	6	4	90	100.0%	HIGH	YES	Pooja_a320	C006-A320
7	C007	Rahul_B737	B737	07JAN2026	7	11	0	50	58.33%	MEDIUM	NO	Rahul_b737	C007-B737
8	C008	Neha_A350	A350	08JAN2026	10	8	2	70	83.33%	HIGH	NO	Neha_a350	C008-A350
9	C009	Vinay_B787	B787	09JAN2026	5	13	0	40	41.67%	LOW	NO	Vinay_b787	C009-B787
10	C010	Kavya_A320	A320	10JAN2026	9	9	1	68	75.00%	MEDIUM	NO	Kavya_a320	C010-A320
11	C011	Ajay_B737	B737	11JAN2026	11	7	3	82	91.67%	HIGH	YES	Ajay_b737	C011-B737
12	C012	Divya_A350	A350	12JAN2026	8	10	1	60	66.67%	MEDIUM	NO	Divya_a350	C012-A350
13	C013	Sanjay_B787	B787	13JAN2026	12	6	4	92	100.0%	HIGH	YES	Sanjay_b787	C013-B787
14	C014	Anu_A320	A320	14JAN2026	6	12	0	48	50.00%	LOW	NO	Anu_a320	C014-A320
15	C015	Rohit_B737	B737	15JAN2026	10	8	2	75	83.33%	HIGH	NO	Rohit_b737	C015-B737
16	C016	Sneha_A350	A350	16JAN2026	7	11	0	55	58.33%	MEDIUM	NO	Sneha_a350	C016-A350

Real data is dirty:

• Extra spaces
• Inconsistent cases
• Poor labeling

Functions clean data for:

• Reporting
• Merging
• Integration

STEP 12 – TRANSPOSE

proc transpose data=crew_schedule out=crew_t;

    var Flight_Hours Rest_Hours Fatigue_Score;

    id Crew_ID;

run;

proc print data=crew_t;

run;

OUTPUT:

Obs	_NAME_	C001	C002	C003	C004	C005	C006	C007	C008	C009	C010	C011	C012	C013	C014	C015	C016
1	Flight_Hours	8	10	6	11	9	12	7	10	5	9	11	8	12	6	10	7
2	Rest_Hours	10	8	12	7	9	6	11	8	13	9	7	10	6	12	8	11
3	Fatigue_Score	60	75	45	85	65	90	50	70	40	68	82	60	92	48	75	55

Used for:

• Excel exports
• Regulatory forms
• Cross-sectional layouts

STEP 13 – APPEND

Creating New Data

data crew_newday;

    input Crew_ID $ Name $13. Aircraft_Type $ Duty_Date :date9.

          Flight_Hours Rest_Hours Delay_Incidents Fatigue_Score;

          Utilization_Pct = Flight_Hours / 12;

    format Duty_Date date9. Utilization_Pct percent8.2;

    datalines;

C017 Aman_A320 A320 17JAN2026 9 9 1 67

C018 Riya_B737 B737 17JAN2026 11 7 3 83

C019 Mohan_A350 A350 17JAN2026 8 10 1 60

C020 Tara_B787 B787 17JAN2026 12 6 4 91

;

run;

proc print data=crew_newday;

run;

OUTPUT:

Obs	Crew_ID	Name	Aircraft_Type	Duty_Date	Flight_Hours	Rest_Hours	Delay_Incidents	Fatigue_Score	Utilization_Pct
1	C017	Aman_A320 A32	0	17JAN2026	9	9	1	67	75.00%
2	C018	Riya_B737 B73	7	17JAN2026	11	7	3	83	91.67%
3	C019	Mohan_A350 A3	50	17JAN2026	8	10	1	60	66.67%
4	C020	Tara_B787 B78	7	17JAN2026	12	6	4	91	100.0%

proc append base=crew_schedule 

            data=crew_newday force;

run;

proc print data=crew_schedule;

run;

OUTPUT:

Obs	Crew_ID	Name	Aircraft_Type	Duty_Date	Flight_Hours	Rest_Hours	Delay_Incidents	Fatigue_Score	Utilization_Pct	Utilization_Status	Fraud_Flag
1	C001	Ramesh_A320	A320	01JAN2026	8	10	1	60	66.67%	MEDIUM	NO
2	C002	Sita_B737	B737	02JAN2026	10	8	2	75	83.33%	HIGH	NO
3	C003	Arjun_A320	A320	03JAN2026	6	12	0	45	50.00%	LOW	NO
4	C004	Meera_A350	A350	04JAN2026	11	7	3	85	91.67%	HIGH	YES
5	C005	Kiran_B787	B787	05JAN2026	9	9	1	65	75.00%	MEDIUM	NO
6	C006	Pooja_A320	A320	06JAN2026	12	6	4	90	100.0%	HIGH	YES
7	C007	Rahul_B737	B737	07JAN2026	7	11	0	50	58.33%	MEDIUM	NO
8	C008	Neha_A350	A350	08JAN2026	10	8	2	70	83.33%	HIGH	NO
9	C009	Vinay_B787	B787	09JAN2026	5	13	0	40	41.67%	LOW	NO
10	C010	Kavya_A320	A320	10JAN2026	9	9	1	68	75.00%	MEDIUM	NO
11	C011	Ajay_B737	B737	11JAN2026	11	7	3	82	91.67%	HIGH	YES
12	C012	Divya_A350	A350	12JAN2026	8	10	1	60	66.67%	MEDIUM	NO
13	C013	Sanjay_B787	B787	13JAN2026	12	6	4	92	100.0%	HIGH	YES
14	C014	Anu_A320	A320	14JAN2026	6	12	0	48	50.00%	LOW	NO
15	C015	Rohit_B737	B737	15JAN2026	10	8	2	75	83.33%	HIGH	NO
16	C016	Sneha_A350	A350	16JAN2026	7	11	0	55	58.33%	MEDIUM	NO
17	C017	Aman_A320 A32	0	17JAN2026	9	9	1	67	75.00%
18	C018	Riya_B737 B73	7	17JAN2026	11	7	3	83	91.67%
19	C019	Mohan_A350 A3	50	17JAN2026	8	10	1	60	66.67%
20	C020	Tara_B787 B78	7	17JAN2026	12	6	4	91	100.0%

Simulates daily data inflow.

Airlines get new records every day.

STEP 14 – SORT & MERGE

proc sort data=crew_schedule;by Aircraft_Type;run;

proc print data=crew_schedule;

run;

OUTPUT:

Obs	Crew_ID	Name	Aircraft_Type	Duty_Date	Flight_Hours	Rest_Hours	Delay_Incidents	Fatigue_Score	Utilization_Pct	Utilization_Status	Fraud_Flag
1	C017	Aman_A320 A32	0	17JAN2026	9	9	1	67	75.00%
2	C019	Mohan_A350 A3	50	17JAN2026	8	10	1	60	66.67%
3	C018	Riya_B737 B73	7	17JAN2026	11	7	3	83	91.67%
4	C020	Tara_B787 B78	7	17JAN2026	12	6	4	91	100.0%
5	C001	Ramesh_A320	A320	01JAN2026	8	10	1	60	66.67%	MEDIUM	NO
6	C003	Arjun_A320	A320	03JAN2026	6	12	0	45	50.00%	LOW	NO
7	C006	Pooja_A320	A320	06JAN2026	12	6	4	90	100.0%	HIGH	YES
8	C010	Kavya_A320	A320	10JAN2026	9	9	1	68	75.00%	MEDIUM	NO
9	C014	Anu_A320	A320	14JAN2026	6	12	0	48	50.00%	LOW	NO
10	C004	Meera_A350	A350	04JAN2026	11	7	3	85	91.67%	HIGH	YES
11	C008	Neha_A350	A350	08JAN2026	10	8	2	70	83.33%	HIGH	NO
12	C012	Divya_A350	A350	12JAN2026	8	10	1	60	66.67%	MEDIUM	NO
13	C016	Sneha_A350	A350	16JAN2026	7	11	0	55	58.33%	MEDIUM	NO
14	C002	Sita_B737	B737	02JAN2026	10	8	2	75	83.33%	HIGH	NO
15	C007	Rahul_B737	B737	07JAN2026	7	11	0	50	58.33%	MEDIUM	NO
16	C011	Ajay_B737	B737	11JAN2026	11	7	3	82	91.67%	HIGH	YES
17	C015	Rohit_B737	B737	15JAN2026	10	8	2	75	83.33%	HIGH	NO
18	C005	Kiran_B787	B787	05JAN2026	9	9	1	65	75.00%	MEDIUM	NO
19	C009	Vinay_B787	B787	09JAN2026	5	13	0	40	41.67%	LOW	NO
20	C013	Sanjay_B787	B787	13JAN2026	12	6	4	92	100.0%	HIGH	YES

proc sort data=summary_sql;by Aircraft_Type;run;

proc print data=summary_sql;

run;

OUTPUT:

Obs	Aircraft_Type	Avg_Flight	Avg_Rest	Total_Delays
1	A320	8.20000	9.80000	6
2	A350	9.00000	9.00000	6
3	B737	9.50000	8.50000	7
4	B787	8.66667	9.33333	5

data merged_data;

    merge crew_schedule 

          summary_sql;

    by Aircraft_Type;

run;

proc print data=merged_data;

run;

OUTPUT:

Obs	Crew_ID	Name	Aircraft_Type	Duty_Date	Flight_Hours	Rest_Hours	Delay_Incidents	Fatigue_Score	Utilization_Pct	Utilization_Status	Fraud_Flag	Avg_Flight	Avg_Rest	Total_Delays
1	C017	Aman_A320 A32	0	17JAN2026	9	9	1	67	75.00%			.	.	.
2	C019	Mohan_A350 A3	50	17JAN2026	8	10	1	60	66.67%			.	.	.
3	C018	Riya_B737 B73	7	17JAN2026	11	7	3	83	91.67%			.	.	.
4	C020	Tara_B787 B78	7	17JAN2026	12	6	4	91	100.0%			.	.	.
5	C001	Ramesh_A320	A320	01JAN2026	8	10	1	60	66.67%	MEDIUM	NO	8.20000	9.80000	6
6	C003	Arjun_A320	A320	03JAN2026	6	12	0	45	50.00%	LOW	NO	8.20000	9.80000	6
7	C006	Pooja_A320	A320	06JAN2026	12	6	4	90	100.0%	HIGH	YES	8.20000	9.80000	6
8	C010	Kavya_A320	A320	10JAN2026	9	9	1	68	75.00%	MEDIUM	NO	8.20000	9.80000	6
9	C014	Anu_A320	A320	14JAN2026	6	12	0	48	50.00%	LOW	NO	8.20000	9.80000	6
10	C004	Meera_A350	A350	04JAN2026	11	7	3	85	91.67%	HIGH	YES	9.00000	9.00000	6
11	C008	Neha_A350	A350	08JAN2026	10	8	2	70	83.33%	HIGH	NO	9.00000	9.00000	6
12	C012	Divya_A350	A350	12JAN2026	8	10	1	60	66.67%	MEDIUM	NO	9.00000	9.00000	6
13	C016	Sneha_A350	A350	16JAN2026	7	11	0	55	58.33%	MEDIUM	NO	9.00000	9.00000	6
14	C002	Sita_B737	B737	02JAN2026	10	8	2	75	83.33%	HIGH	NO	9.50000	8.50000	7
15	C007	Rahul_B737	B737	07JAN2026	7	11	0	50	58.33%	MEDIUM	NO	9.50000	8.50000	7
16	C011	Ajay_B737	B737	11JAN2026	11	7	3	82	91.67%	HIGH	YES	9.50000	8.50000	7
17	C015	Rohit_B737	B737	15JAN2026	10	8	2	75	83.33%	HIGH	NO	9.50000	8.50000	7
18	C005	Kiran_B787	B787	05JAN2026	9	9	1	65	75.00%	MEDIUM	NO	8.66667	9.33333	5
19	C009	Vinay_B787	B787	09JAN2026	5	13	0	40	41.67%	LOW	NO	8.66667	9.33333	5
20	C013	Sanjay_B787	B787	13JAN2026	12	6	4	92	100.0%	HIGH	YES	8.66667	9.33333	5

Combines operational data with:

• Summaries
• Benchmarks
• Targets

STEP 15 – DELETE USING PROC DATASETS

proc datasets library=work;

    delete char_demo crew_t;

quit;

LOG:

NOTE: Deleting WORK.CHAR_DEMO (memtype=DATA).

NOTE: Deleting WORK.CREW_T (memtype=DATA).

Professional hygiene:

• Free memory
• Avoid confusion
• Speed processing

THIS PROJECT COVERS

This single project prepares you for:

Topic	Covered
Base SAS	YES
SQL	YES
Macros	YES
Dates	YES
Character functions	YES
Numeric functions	YES
Reporting	YES
Visualization	YES
Data integration	YES
Business logic	YES

FINAL BUSINESS IMPACT

If this system runs in real life:

Airline gets:

·       Safer operations

·       Lower accident risk

·       Better crew morale

·       Lower regulatory penalties

·       Higher efficiency

This is exactly what aviation analytics teams do daily.

FUTURE ENHANCEMENTS

This system can be expanded with:

·       Weather impact analysis

·       Time zone adjustments

·       Multi-leg flight modeling

·       Predictive machine learning

·       Real-time streaming data

·       Integration with HR systems

CONCLUSION

The Airline Crew Schedules Analytics System demonstrates how SAS can be used as a complete operational intelligence platform in a highly regulated, safety-critical industry.

This project is not a simple academic exercise. It reflects how real airlines:

·       Collect operational data

·       Apply business rules

·       Detect risks

·       Visualize patterns

·       Support decision making

·       Ensure regulatory compliance

By using:

·       PROC SQL for summarization

·       PROC MEANS and UNIVARIATE for statistics

·       PROC CORR for relationship analysis

·       PROC SGPLOT for visualization

·       SAS Macros for automation

·       Date functions for scheduling logic

·       Character and numeric functions for data quality

This system becomes a fully functional airline analytics framework.

INTERVIEW QUESTIONS FOR YOU

·  What is the difference between CAT, CATS, and CATX?

·  What is the difference between TRIM and STRIP?

·  What is SCAN function?

 

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

About the Author:

SAS Learning Hub is a data analytics and SAS programming platform focused on clinical, financial, and real-world data analysis. The content is created by professionals with academic training in Pharmaceutics and hands-on experience in Base SAS, PROC SQL, Macros, SDTM, and ADaM, providing practical and industry-relevant SAS learning resources.

Disclaimer:

The datasets and analysis in this article are created for educational and demonstration purposes only. They do not represent Airline Crew data.

Our Mission:

This blog provides industry-focused SAS programming tutorials and analytics projects covering finance, healthcare, and technology.

This project is suitable for:

·  Students learning SAS

·  Data analysts building portfolios

·  Professionals preparing for SAS interviews

·  Bloggers writing about analytics and smart cities

·  EV and energy industry professionals

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

Follow Us On : 

 

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

--->Follow our blog for more SAS-based analytics projects and industry data models.

---> Support Us By Following Our Blog..

To deepen your understanding of SAS analytics, please refer to our other data science and industry-focused projects listed below:

1.Can we analyze planets using SAS like NASA scientists?

2.Can SAS Compare How 3 Different Families Live in Their Homes?

3.Can SAS Help Us Choose the Most Efficient Cooler?

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

About Us | Contact | Privacy Policy