405.What Do Vehicles, Waiting Time, and Cashless Payments Really Say About Toll Plaza Performance?A Complete SAS Analytics Investigation

What Happens When High Traffic Meets Long Waiting Time?

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

HERE WE USED THESE SAS STATEMENTS AND FUNCTIONS FOR THIS PROJECT:

DATA STEP | SET | INPUT | DATALINES | LENGTH | FORMAT | INFORMAT | IF-THEN-ELSE | RENAME | DROP | KEEP | LABEL | ROUND | INTNX | INTCK | MDY | STRIP | TRIM | CAT | CATX | PROPCASE | UPCASE | LOWCASE | COALESCE | PROC CONTENTS | PROC SQL | CREATE TABLE | SELECT | GROUP BY | ORDER BY | PROC FREQ | PROC MEANS | PROC UNIVARIATE | PROC CORR | PROC SGPLOT | SCATTER | PROC APPEND | PROC TRANSPOSE | PROC DATASETS DELETE | MERGE | BY | %MACRO | %MEND

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

INTRODUCTION

Toll plazas are critical infrastructure points that directly impact:

·       Traffic flow

·       Revenue collection

·       Public satisfaction

·       Fraud risk

·       Digital payment adoption

This project simulates real-world toll plaza data and demonstrates how a SAS programmer would analyze operations, efficiency, utilization, and fraud patterns using Base SAS, PROC SQL, Macros, and Analytics procedures.

TABLE OF CONTENTS

1.     Business Context

2.     Project Objectives

3.     Dataset Design

4.     Raw Data Creation (with intentional error)

5.     Identifying the Error

6.     Correcting the Error

7.     Final Clean Dataset Creation

8.     Date Handling & Derivations

9.     Efficiency & Utilization Logic (Macro)

10.  Fraud Detection Logic (Macro)

11.  PROC SQL Analytics

12.  PROC FREQ, MEANS, UNIVARIATE

13.  Correlation Analysis

14.  Visualization (PROC SGPLOT)

15.  Append, Set, Merge, Transpose

16.  Character & Numeric Functions

17.  PROC DATASETS Delete

18.  Business Insights

19.  Key Points

20.  Conclusion

1. BUSINESS CONTEXT

Government authorities and private concessionaires monitor toll plazas to:

·       Reduce congestion

·       Improve digital payments (FASTag / RFID)

·       Detect revenue leakage

·       Improve operational efficiency

SAS is widely used in:

·       Transport analytics

·       Government data platforms

·       Fraud and compliance systems

2. PROJECT OBJECTIVES

·       Create a realistic toll plaza dataset

·       Introduce a deliberate data error

·       Detect and correct the error

·       Calculate Efficiency Score (%)

·       Classify Utilization Levels

·       Identify Fraud Risk patterns

·       Use full SAS toolchain

3. DATASET DESIGN

Variables

Variable	Description
Plaza_ID	Unique toll plaza ID
Highway_Name	Name of highway
Vehicles_Per_Day	Average daily vehicles
Revenue	Daily revenue (INR)
Avg_Wait_Time	Avg waiting time (minutes)
Cashless_Ratio	% of digital payments
Efficiency_Score	Calculated (%)
Collection_Date	Observation date
Region	Geographic region
Fraud_Flag	Fraud indicator

 

Plaza_ID

A unique identifier for each toll plaza.
Used as a primary reference across datasets.

Highway_Name

Indicates the highway on which the toll plaza operates.
Useful for infrastructure planning and comparative studies.

Vehicles_Per_Day

Represents average daily traffic volume.
This is a core operational KPI.

Revenue

Daily revenue collected at the toll plaza.
Highly sensitive and audit-critical.

Avg_Wait_Time

Average waiting time per vehicle (minutes).
Directly impacts public satisfaction.

Cashless_Ratio

Percentage of vehicles using digital payment methods.
Indicates modernization and efficiency.

Efficiency_Score

Derived metric measuring operational effectiveness.
Combines traffic, payment mode, and waiting time.

Collection_Date

Date of data collection.
Enables trend and time-series analysis.

Region

Geographical grouping for regional insights.

Fraud_Flag

Derived indicator highlighting suspicious patterns.

 

4. RAW DATA CREATION (WITH INTENTIONAL ERROR)

Intentional Error:

Revenue typed as character ($) instead of numeric

This will break calculations later

data toll_raw;

    length Plaza_ID $6 Highway_Name $30 Region $15 Revenue $12;

    input Plaza_ID $ Highway_Name $ Vehicles_Per_Day Revenue $ Avg_Wait_Time

          Cashless_Ratio Collection_Date : date9. Region $;

    format Collection_Date date9.;

datalines;

TP001 NH44 45000 850000 3.5 82 01JAN2026 South

TP002 NH48 38000 720000 4.2 78 01JAN2026 West

TP003 NH16 51000 910000 2.9 88 01JAN2026 East

TP004 NH65 29000 560000 6.8 60 01JAN2026 South

TP005 NH19 47000 880000 3.1 85 01JAN2026 North

TP006 NH27 26000 490000 7.5 55 01JAN2026 West

TP007 NH30 34000 650000 5.2 70 01JAN2026 Central

TP008 NH66 52000 970000 2.7 90 01JAN2026 South

TP009 NH52 31000 600000 6.1 62 01JAN2026 North

TP010 NH40 28000 540000 7.0 58 01JAN2026 East

TP011 NH75 36000 690000 4.8 73 01JAN2026 West

TP012 NH22 24000 460000 8.0 50 01JAN2026 North

TP013 NH13 49000 920000 3.0 86 01JAN2026 South

;

run;

proc print data=toll_raw;

run;

OUTPUT:

Obs	Plaza_ID	Highway_Name	Region	Revenue	Vehicles_Per_Day	Avg_Wait_Time	Cashless_Ratio	Collection_Date
1	TP001	NH44	South	850000	45000	3.5	82	01JAN2026
2	TP002	NH48	West	720000	38000	4.2	78	01JAN2026
3	TP003	NH16	East	910000	51000	2.9	88	01JAN2026
4	TP004	NH65	South	560000	29000	6.8	60	01JAN2026
5	TP005	NH19	North	880000	47000	3.1	85	01JAN2026
6	TP006	NH27	West	490000	26000	7.5	55	01JAN2026
7	TP007	NH30	Central	650000	34000	5.2	70	01JAN2026
8	TP008	NH66	South	970000	52000	2.7	90	01JAN2026
9	TP009	NH52	North	600000	31000	6.1	62	01JAN2026
10	TP010	NH40	East	540000	28000	7.0	58	01JAN2026
11	TP011	NH75	West	690000	36000	4.8	73	01JAN2026
12	TP012	NH22	North	460000	24000	8.0	50	01JAN2026
13	TP013	NH13	South	920000	49000	3.0	86	01JAN2026

Common issues:

·       Numeric values stored as character

·       Missing values

·       Format mismatches

·       Inconsistent lengths

This project intentionally introduces an error where Revenue is stored as a character variable instead of numeric.

This simulates:

·       Manual data entry mistakes

·       Poor upstream system design

·       CSV or Excel ingestion issues

5. IDENTIFYING THE ERROR

proc contents data=toll_raw;

run;

OUTPUT:

The CONTENTS Procedure

Data Set Name	WORK.TOLL_RAW	Observations	13
Member Type	DATA	Variables	8
Engine	V9	Indexes	0
Created	02/18/2026 19:27:00	Observation Length	96
Last Modified	02/18/2026 19:27:00	Deleted Observations	0
Protection		Compressed	NO
Data Set Type		Sorted	NO
Label
Data Representation	SOLARIS_X86_64, LINUX_X86_64, ALPHA_TRU64, LINUX_IA64
Encoding	utf-8 Unicode (UTF-8)

Engine/Host Dependent Information
Data Set Page Size	131072
Number of Data Set Pages	1
First Data Page	1
Max Obs per Page	1363
Obs in First Data Page	13
Number of Data Set Repairs	0
Filename	/saswork/SAS_work9BFC00019166_odaws02-apse1-2.oda.sas.com/SAS_workD59100019166_odaws02-apse1-2.oda.sas.com/toll_raw.sas7bdat
Release Created	9.0401M8
Host Created	Linux
Inode Number	201335737
Access Permission	rw-r--r--
Owner Name	u63247146
File Size	256KB
File Size (bytes)	262144

Alphabetic List of Variables and Attributes
#	Variable	Type	Len	Format
6	Avg_Wait_Time	Num	8
7	Cashless_Ratio	Num	8
8	Collection_Date	Num	8	DATE9.
2	Highway_Name	Char	30
1	Plaza_ID	Char	6
3	Region	Char	15
4	Revenue	Char	12
5	Vehicles_Per_Day	Num	8

PROC CONTENTS is one of the most underrated but critical procedures in SAS.

It allows us to:

·  Validate variable types

·  Check lengths and formats

·  Confirm dataset structure

Problem Identified

·  Revenue is character

·  Cannot compute efficiency or fraud metrics

If Revenue is character:

·  Arithmetic operations fail

·  Efficiency calculations break

·  Fraud logic becomes invalid

·  Reports become misleading

6. CORRECTING THE ERROR

data toll_clean;

    set toll_raw;

    Revenue_Num = input(strip(Revenue), best12.);

    drop Revenue;

    rename Revenue_Num = Revenue;

run;

proc print data=toll_clean;

run;

OUTPUT:

Obs	Plaza_ID	Highway_Name	Region	Vehicles_Per_Day	Avg_Wait_Time	Cashless_Ratio	Collection_Date	Revenue
1	TP001	NH44	South	45000	3.5	82	01JAN2026	850000
2	TP002	NH48	West	38000	4.2	78	01JAN2026	720000
3	TP003	NH16	East	51000	2.9	88	01JAN2026	910000
4	TP004	NH65	South	29000	6.8	60	01JAN2026	560000
5	TP005	NH19	North	47000	3.1	85	01JAN2026	880000
6	TP006	NH27	West	26000	7.5	55	01JAN2026	490000
7	TP007	NH30	Central	34000	5.2	70	01JAN2026	650000
8	TP008	NH66	South	52000	2.7	90	01JAN2026	970000
9	TP009	NH52	North	31000	6.1	62	01JAN2026	600000
10	TP010	NH40	East	28000	7.0	58	01JAN2026	540000
11	TP011	NH75	West	36000	4.8	73	01JAN2026	690000
12	TP012	NH22	North	24000	8.0	50	01JAN2026	460000
13	TP013	NH13	South	49000	3.0	86	01JAN2026	920000

·  STRIP ( ) removes spaces

·  INPUT ( ) converts character → numeric

·  Dataset becomes analytics-ready

7. FINAL CLEAN DATASET

data toll_final;

    set toll_clean;

    Efficiency_Score =

        round((Cashless_Ratio * Vehicles_Per_Day) /(Avg_Wait_Time * 1000),0.01);

    format Efficiency_Score percent8.2;

run;

proc print data=toll_final;

run;

OUTPUT:

Obs	Plaza_ID	Highway_Name	Region	Vehicles_Per_Day	Avg_Wait_Time	Cashless_Ratio	Collection_Date	Revenue	Efficiency_Score
1	TP001	NH44	South	45000	3.5	82	01JAN2026	850000	105E3%
2	TP002	NH48	West	38000	4.2	78	01JAN2026	720000	70571%
3	TP003	NH16	East	51000	2.9	88	01JAN2026	910000	155E3%
4	TP004	NH65	South	29000	6.8	60	01JAN2026	560000	25588%
5	TP005	NH19	North	47000	3.1	85	01JAN2026	880000	129E3%
6	TP006	NH27	West	26000	7.5	55	01JAN2026	490000	19067%
7	TP007	NH30	Central	34000	5.2	70	01JAN2026	650000	45769%
8	TP008	NH66	South	52000	2.7	90	01JAN2026	970000	173E3%
9	TP009	NH52	North	31000	6.1	62	01JAN2026	600000	31508%
10	TP010	NH40	East	28000	7.0	58	01JAN2026	540000	23200%
11	TP011	NH75	West	36000	4.8	73	01JAN2026	690000	54750%
12	TP012	NH22	North	24000	8.0	50	01JAN2026	460000	15000%
13	TP013	NH13	South	49000	3.0	86	01JAN2026	920000	14E4%

WHY PERCENT FORMAT IS USED:

Efficiency is expressed as a percentage because:

·  It is easy to compare across plazas

·  It is intuitive for business users

·  It aligns with KPI dashboards

8. DATE HANDLING & DERIVATIONS

data toll_dates;

    set toll_final;

    Month = month(Collection_Date);

    Next_Month = intnx('month', Collection_Date, 1);

    Days_Since = intck('day', Collection_Date, today());

    format Next_Month date9.;

run;

proc print data=toll_dates;

run;

OUTPUT:

Obs	Plaza_ID	Highway_Name	Region	Vehicles_Per_Day	Avg_Wait_Time	Cashless_Ratio	Collection_Date	Revenue	Efficiency_Score	Month	Next_Month	Days_Since
1	TP001	NH44	South	45000	3.5	82	01JAN2026	850000	105E3%	1	01FEB2026	48
2	TP002	NH48	West	38000	4.2	78	01JAN2026	720000	70571%	1	01FEB2026	48
3	TP003	NH16	East	51000	2.9	88	01JAN2026	910000	155E3%	1	01FEB2026	48
4	TP004	NH65	South	29000	6.8	60	01JAN2026	560000	25588%	1	01FEB2026	48
5	TP005	NH19	North	47000	3.1	85	01JAN2026	880000	129E3%	1	01FEB2026	48
6	TP006	NH27	West	26000	7.5	55	01JAN2026	490000	19067%	1	01FEB2026	48
7	TP007	NH30	Central	34000	5.2	70	01JAN2026	650000	45769%	1	01FEB2026	48
8	TP008	NH66	South	52000	2.7	90	01JAN2026	970000	173E3%	1	01FEB2026	48
9	TP009	NH52	North	31000	6.1	62	01JAN2026	600000	31508%	1	01FEB2026	48
10	TP010	NH40	East	28000	7.0	58	01JAN2026	540000	23200%	1	01FEB2026	48
11	TP011	NH75	West	36000	4.8	73	01JAN2026	690000	54750%	1	01FEB2026	48
12	TP012	NH22	North	24000	8.0	50	01JAN2026	460000	15000%	1	01FEB2026	48
13	TP013	NH13	South	49000	3.0	86	01JAN2026	920000	14E4%	1	01FEB2026	48

Functions used:

·  INTCK() → difference between dates

·  INTNX() → shifting dates

In toll systems, fraud indicators include:

·  High traffic but low revenue

·  Sudden revenue drops

·  Inconsistent payment ratios

9. UTILIZATION CLASSIFICATION MACRO

%macro utilization;

data toll_util;

    set toll_dates;

    length Utilization $10.;

    if Vehicles_Per_Day >= 50000 then Utilization = 'Very High';

    else if Vehicles_Per_Day >= 35000 then Utilization = 'High';

    else if Vehicles_Per_Day >= 25000 then Utilization = 'Medium';

    else Utilization = 'Low';

run;

proc print data=toll_util;

run;

%mend;

%utilization;

OUTPUT:

Obs	Plaza_ID	Highway_Name	Region	Vehicles_Per_Day	Avg_Wait_Time	Cashless_Ratio	Collection_Date	Revenue	Efficiency_Score	Month	Next_Month	Days_Since	Utilization
1	TP001	NH44	South	45000	3.5	82	01JAN2026	850000	105E3%	1	01FEB2026	48	High
2	TP002	NH48	West	38000	4.2	78	01JAN2026	720000	70571%	1	01FEB2026	48	High
3	TP003	NH16	East	51000	2.9	88	01JAN2026	910000	155E3%	1	01FEB2026	48	Very High
4	TP004	NH65	South	29000	6.8	60	01JAN2026	560000	25588%	1	01FEB2026	48	Medium
5	TP005	NH19	North	47000	3.1	85	01JAN2026	880000	129E3%	1	01FEB2026	48	High
6	TP006	NH27	West	26000	7.5	55	01JAN2026	490000	19067%	1	01FEB2026	48	Medium
7	TP007	NH30	Central	34000	5.2	70	01JAN2026	650000	45769%	1	01FEB2026	48	Medium
8	TP008	NH66	South	52000	2.7	90	01JAN2026	970000	173E3%	1	01FEB2026	48	Very High
9	TP009	NH52	North	31000	6.1	62	01JAN2026	600000	31508%	1	01FEB2026	48	Medium
10	TP010	NH40	East	28000	7.0	58	01JAN2026	540000	23200%	1	01FEB2026	48	Medium
11	TP011	NH75	West	36000	4.8	73	01JAN2026	690000	54750%	1	01FEB2026	48	High
12	TP012	NH22	North	24000	8.0	50	01JAN2026	460000	15000%	1	01FEB2026	48	Low
13	TP013	NH13	South	49000	3.0	86	01JAN2026	920000	14E4%	1	01FEB2026	48	High

 

10. FRAUD DETECTION MACRO

%macro fraud_check;

data toll_fraud;

    set toll_util;

    length Fraud_Flag $3.;

    if Revenue < 500000 and Vehicles_Per_Day > 20000 then Fraud_Flag = 'YES';

    else Fraud_Flag = 'NO';

run;

proc print data=toll_fraud;

run;

%mend;

%fraud_check;

OUTPUT:

Obs	Plaza_ID	Highway_Name	Region	Vehicles_Per_Day	Avg_Wait_Time	Cashless_Ratio	Collection_Date	Revenue	Efficiency_Score	Month	Next_Month	Days_Since	Utilization	Fraud_Flag
1	TP001	NH44	South	45000	3.5	82	01JAN2026	850000	105E3%	1	01FEB2026	48	High	NO
2	TP002	NH48	West	38000	4.2	78	01JAN2026	720000	70571%	1	01FEB2026	48	High	NO
3	TP003	NH16	East	51000	2.9	88	01JAN2026	910000	155E3%	1	01FEB2026	48	Very High	NO
4	TP004	NH65	South	29000	6.8	60	01JAN2026	560000	25588%	1	01FEB2026	48	Medium	NO
5	TP005	NH19	North	47000	3.1	85	01JAN2026	880000	129E3%	1	01FEB2026	48	High	NO
6	TP006	NH27	West	26000	7.5	55	01JAN2026	490000	19067%	1	01FEB2026	48	Medium	YES
7	TP007	NH30	Central	34000	5.2	70	01JAN2026	650000	45769%	1	01FEB2026	48	Medium	NO
8	TP008	NH66	South	52000	2.7	90	01JAN2026	970000	173E3%	1	01FEB2026	48	Very High	NO
9	TP009	NH52	North	31000	6.1	62	01JAN2026	600000	31508%	1	01FEB2026	48	Medium	NO
10	TP010	NH40	East	28000	7.0	58	01JAN2026	540000	23200%	1	01FEB2026	48	Medium	NO
11	TP011	NH75	West	36000	4.8	73	01JAN2026	690000	54750%	1	01FEB2026	48	High	NO
12	TP012	NH22	North	24000	8.0	50	01JAN2026	460000	15000%	1	01FEB2026	48	Low	YES
13	TP013	NH13	South	49000	3.0	86	01JAN2026	920000	14E4%	1	01FEB2026	48	High	NO

In toll systems, fraud indicators include:

·  High traffic but low revenue

·  Sudden revenue drops

·  Inconsistent payment ratios

11. PROC SQL ANALYTICS

proc sql;

    create table region_summary as

    select Region,

           count(*) as Plaza_Count,

           avg(Revenue) as Avg_Revenue format=comma12.,

           avg(Efficiency_Score) as Avg_Efficiency

    from toll_fraud

    group by Region;

quit;

proc print data=region_summary;

run;

OUTPUT:

Obs	Region	Plaza_Count	Avg_Revenue	Avg_Efficiency
1	Central	1	650,000	457.69
2	East	2	725,000	889.80
3	North	3	646,667	584.60
4	South	4	825,000	1112.04
5	West	3	633,333	481.29

·  It is powerful for aggregation

·  Business users understand SQL logic

·  It simplifies grouping and summarization

12. PROC FREQ, MEANS, UNIVARIATE

proc freq data=toll_fraud;

    tables Utilization Fraud_Flag;

run;

OUTPUT:

The FREQ Procedure

Utilization	Frequency	Percent	Cumulative Frequency	Cumulative Percent
High	5	38.46	5	38.46
Low	1	7.69	6	46.15
Medium	5	38.46	11	84.62
Very High	2	15.38	13	100.00

Fraud_Flag	Frequency	Percent	Cumulative Frequency	Cumulative Percent
NO	11	84.62	11	84.62
YES	2	15.38	13	100.00

proc means data=toll_fraud mean min max;

    var Revenue Vehicles_Per_Day Avg_Wait_Time;

run;

OUTPUT:

The MEANS Procedure

Variable	Mean	Minimum	Maximum
Revenue Vehicles_Per_Day Avg_Wait_Time	710769.23 37692.31 4.9846154	460000.00 24000.00 2.7000000	970000.00 52000.00 8.0000000

PROC MEANS

·  Quick summary

·  Operational monitoring

proc univariate data=toll_fraud;

    var Revenue;

run;

OUTPUT:

The UNIVARIATE Procedure

Variable: Revenue

Moments
N	13	Sum Weights	13
Mean	710769.231	Sum Observations	9240000
Std Deviation	177644.849	Variance	3.15577E10
Skewness	0.096711	Kurtosis	-1.5645914
Uncorrected SS	6.9462E12	Corrected SS	3.78692E11
Coeff Variation	24.9933229	Std Error Mean	49269.8162

Basic Statistical Measures
Location		Variability
Mean	710769.2	Std Deviation	177645
Median	690000.0	Variance	3.15577E10
Mode	.	Range	510000
		Interquartile Range	320000

Tests for Location: Mu0=0
Test	Statistic		p Value
Student's t	t	14.42606	Pr > |t|	<.0001
Sign	M	6.5	Pr >= |M|	0.0002
Signed Rank	S	45.5	Pr >= |S|	0.0002

Quantiles (Definition 5)
Level	Quantile
100% Max	970000
99%	970000
95%	970000
90%	920000
75% Q3	880000
50% Median	690000
25% Q1	560000
10%	490000
5%	460000
1%	460000
0% Min	460000

Extreme Observations
Lowest		Highest
Value	Obs	Value	Obs
460000	12	850000	1
490000	6	880000	5
540000	10	910000	3
560000	4	920000	13
600000	9	970000	8

PROC UNIVARIATE

·  Detailed distribution

·  Outlier detection

·  Statistical depth

13. CORRELATION ANALYSIS

proc corr data=toll_fraud;

    var Vehicles_Per_Day Revenue Cashless_Ratio Avg_Wait_Time;

run;

OUTPUT:

The CORR Procedure

4 Variables:	Vehicles_Per_Day Revenue Cashless_Ratio Avg_Wait_Time

Simple Statistics
Variable	N	Mean	Std Dev	Sum	Minimum	Maximum
Vehicles_Per_Day	13	37692	10020	490000	24000	52000
Revenue	13	710769	177645	9240000	460000	970000
Cashless_Ratio	13	72.07692	13.84113	937.00000	50.00000	90.00000
Avg_Wait_Time	13	4.98462	1.91435	64.80000	2.70000	8.00000

Pearson Correlation Coefficients, N = 13 Prob > |r| under H0: Rho=0
	Vehicles_Per_Day	Revenue	Cashless_Ratio	Avg_Wait_Time
Vehicles_Per_Day	1.00000	0.99735 <.0001	0.98382 <.0001	-0.97951 <.0001
Revenue	0.99735 <.0001	1.00000	0.98656 <.0001	-0.98455 <.0001
Cashless_Ratio	0.98382 <.0001	0.98656 <.0001	1.00000	-0.99756 <.0001
Avg_Wait_Time	-0.97951 <.0001	-0.98455 <.0001	-0.99756 <.0001	1.00000

14. VISUALIZATION

proc sgplot data=toll_fraud;

    scatter x=Vehicles_Per_Day y=Revenue;

    title "Vehicles vs Revenue";

run;

OUTPUT:

15. APPEND,TRANSPOSE

proc append base=toll_fraud 

            data=toll_fraud force;

run;

proc print data=toll_fraud;

run;

OUTPUT:

Obs	Plaza_ID	Highway_Name	Region	Vehicles_Per_Day	Avg_Wait_Time	Cashless_Ratio	Collection_Date	Revenue	Efficiency_Score	Month	Next_Month	Days_Since	Utilization	Fraud_Flag
1	TP001	NH44	South	45000	3.5	82	01JAN2026	850000	105E3%	1	01FEB2026	48	High	NO
2	TP002	NH48	West	38000	4.2	78	01JAN2026	720000	70571%	1	01FEB2026	48	High	NO
3	TP003	NH16	East	51000	2.9	88	01JAN2026	910000	155E3%	1	01FEB2026	48	Very High	NO
4	TP004	NH65	South	29000	6.8	60	01JAN2026	560000	25588%	1	01FEB2026	48	Medium	NO
5	TP005	NH19	North	47000	3.1	85	01JAN2026	880000	129E3%	1	01FEB2026	48	High	NO
6	TP006	NH27	West	26000	7.5	55	01JAN2026	490000	19067%	1	01FEB2026	48	Medium	YES
7	TP007	NH30	Central	34000	5.2	70	01JAN2026	650000	45769%	1	01FEB2026	48	Medium	NO
8	TP008	NH66	South	52000	2.7	90	01JAN2026	970000	173E3%	1	01FEB2026	48	Very High	NO
9	TP009	NH52	North	31000	6.1	62	01JAN2026	600000	31508%	1	01FEB2026	48	Medium	NO
10	TP010	NH40	East	28000	7.0	58	01JAN2026	540000	23200%	1	01FEB2026	48	Medium	NO
11	TP011	NH75	West	36000	4.8	73	01JAN2026	690000	54750%	1	01FEB2026	48	High	NO
12	TP012	NH22	North	24000	8.0	50	01JAN2026	460000	15000%	1	01FEB2026	48	Low	YES
13	TP013	NH13	South	49000	3.0	86	01JAN2026	920000	14E4%	1	01FEB2026	48	High	NO
14	TP001	NH44	South	45000	3.5	82	01JAN2026	850000	105E3%	1	01FEB2026	48	High	NO
15	TP002	NH48	West	38000	4.2	78	01JAN2026	720000	70571%	1	01FEB2026	48	High	NO
16	TP003	NH16	East	51000	2.9	88	01JAN2026	910000	155E3%	1	01FEB2026	48	Very High	NO
17	TP004	NH65	South	29000	6.8	60	01JAN2026	560000	25588%	1	01FEB2026	48	Medium	NO
18	TP005	NH19	North	47000	3.1	85	01JAN2026	880000	129E3%	1	01FEB2026	48	High	NO
19	TP006	NH27	West	26000	7.5	55	01JAN2026	490000	19067%	1	01FEB2026	48	Medium	YES
20	TP007	NH30	Central	34000	5.2	70	01JAN2026	650000	45769%	1	01FEB2026	48	Medium	NO
21	TP008	NH66	South	52000	2.7	90	01JAN2026	970000	173E3%	1	01FEB2026	48	Very High	NO
22	TP009	NH52	North	31000	6.1	62	01JAN2026	600000	31508%	1	01FEB2026	48	Medium	NO
23	TP010	NH40	East	28000	7.0	58	01JAN2026	540000	23200%	1	01FEB2026	48	Medium	NO
24	TP011	NH75	West	36000	4.8	73	01JAN2026	690000	54750%	1	01FEB2026	48	High	NO
25	TP012	NH22	North	24000	8.0	50	01JAN2026	460000	15000%	1	01FEB2026	48	Low	YES
26	TP013	NH13	South	49000	3.0	86	01JAN2026	920000	14E4%	1	01FEB2026	48	High	NO

What Is Happening Here?

we are trying to append a dataset to itself.

·       BASE= → toll_fraud

·       DATA= → toll_fraud

·       Both point to the same physical dataset

Step 1:New Day Toll Data

data toll_fraud_new;

    set toll_fraud;

    Collection_Date = intnx('day', Collection_Date, 1);

    Revenue = Revenue * 1.03;

run;

proc print data=toll_fraud_new;

run;

OUTPUT:

Obs	Plaza_ID	Highway_Name	Region	Vehicles_Per_Day	Avg_Wait_Time	Cashless_Ratio	Collection_Date	Revenue	Efficiency_Score	Month	Next_Month	Days_Since	Utilization	Fraud_Flag
1	TP001	NH44	South	45000	3.5	82	02JAN2026	875500	105E3%	1	01FEB2026	48	High	NO
2	TP002	NH48	West	38000	4.2	78	02JAN2026	741600	70571%	1	01FEB2026	48	High	NO
3	TP003	NH16	East	51000	2.9	88	02JAN2026	937300	155E3%	1	01FEB2026	48	Very High	NO
4	TP004	NH65	South	29000	6.8	60	02JAN2026	576800	25588%	1	01FEB2026	48	Medium	NO
5	TP005	NH19	North	47000	3.1	85	02JAN2026	906400	129E3%	1	01FEB2026	48	High	NO
6	TP006	NH27	West	26000	7.5	55	02JAN2026	504700	19067%	1	01FEB2026	48	Medium	YES
7	TP007	NH30	Central	34000	5.2	70	02JAN2026	669500	45769%	1	01FEB2026	48	Medium	NO
8	TP008	NH66	South	52000	2.7	90	02JAN2026	999100	173E3%	1	01FEB2026	48	Very High	NO
9	TP009	NH52	North	31000	6.1	62	02JAN2026	618000	31508%	1	01FEB2026	48	Medium	NO
10	TP010	NH40	East	28000	7.0	58	02JAN2026	556200	23200%	1	01FEB2026	48	Medium	NO
11	TP011	NH75	West	36000	4.8	73	02JAN2026	710700	54750%	1	01FEB2026	48	High	NO
12	TP012	NH22	North	24000	8.0	50	02JAN2026	473800	15000%	1	01FEB2026	48	Low	YES
13	TP013	NH13	South	49000	3.0	86	02JAN2026	947600	14E4%	1	01FEB2026	48	High	NO

Step 2:Correct Append

proc append base=toll_fraud

            data=toll_fraud_new force;

run;

proc print data=toll_fraud;

run;

OUTPUT:

Obs	Plaza_ID	Highway_Name	Region	Vehicles_Per_Day	Avg_Wait_Time	Cashless_Ratio	Collection_Date	Revenue	Efficiency_Score	Month	Next_Month	Days_Since	Utilization	Fraud_Flag
1	TP001	NH44	South	45000	3.5	82	01JAN2026	850000	105E3%	1	01FEB2026	48	High	NO
2	TP002	NH48	West	38000	4.2	78	01JAN2026	720000	70571%	1	01FEB2026	48	High	NO
3	TP003	NH16	East	51000	2.9	88	01JAN2026	910000	155E3%	1	01FEB2026	48	Very High	NO
4	TP004	NH65	South	29000	6.8	60	01JAN2026	560000	25588%	1	01FEB2026	48	Medium	NO
5	TP005	NH19	North	47000	3.1	85	01JAN2026	880000	129E3%	1	01FEB2026	48	High	NO
6	TP006	NH27	West	26000	7.5	55	01JAN2026	490000	19067%	1	01FEB2026	48	Medium	YES
7	TP007	NH30	Central	34000	5.2	70	01JAN2026	650000	45769%	1	01FEB2026	48	Medium	NO
8	TP008	NH66	South	52000	2.7	90	01JAN2026	970000	173E3%	1	01FEB2026	48	Very High	NO
9	TP009	NH52	North	31000	6.1	62	01JAN2026	600000	31508%	1	01FEB2026	48	Medium	NO
10	TP010	NH40	East	28000	7.0	58	01JAN2026	540000	23200%	1	01FEB2026	48	Medium	NO
11	TP011	NH75	West	36000	4.8	73	01JAN2026	690000	54750%	1	01FEB2026	48	High	NO
12	TP012	NH22	North	24000	8.0	50	01JAN2026	460000	15000%	1	01FEB2026	48	Low	YES
13	TP013	NH13	South	49000	3.0	86	01JAN2026	920000	14E4%	1	01FEB2026	48	High	NO
14	TP001	NH44	South	45000	3.5	82	02JAN2026	875500	105E3%	1	01FEB2026	48	High	NO
15	TP002	NH48	West	38000	4.2	78	02JAN2026	741600	70571%	1	01FEB2026	48	High	NO
16	TP003	NH16	East	51000	2.9	88	02JAN2026	937300	155E3%	1	01FEB2026	48	Very High	NO
17	TP004	NH65	South	29000	6.8	60	02JAN2026	576800	25588%	1	01FEB2026	48	Medium	NO
18	TP005	NH19	North	47000	3.1	85	02JAN2026	906400	129E3%	1	01FEB2026	48	High	NO
19	TP006	NH27	West	26000	7.5	55	02JAN2026	504700	19067%	1	01FEB2026	48	Medium	YES
20	TP007	NH30	Central	34000	5.2	70	02JAN2026	669500	45769%	1	01FEB2026	48	Medium	NO
21	TP008	NH66	South	52000	2.7	90	02JAN2026	999100	173E3%	1	01FEB2026	48	Very High	NO
22	TP009	NH52	North	31000	6.1	62	02JAN2026	618000	31508%	1	01FEB2026	48	Medium	NO
23	TP010	NH40	East	28000	7.0	58	02JAN2026	556200	23200%	1	01FEB2026	48	Medium	NO
24	TP011	NH75	West	36000	4.8	73	02JAN2026	710700	54750%	1	01FEB2026	48	High	NO
25	TP012	NH22	North	24000	8.0	50	02JAN2026	473800	15000%	1	01FEB2026	48	Low	YES
26	TP013	NH13	South	49000	3.0	86	02JAN2026	947600	14E4%	1	01FEB2026	48	High	NO

PROC APPEND is designed for:

·  Daily incremental loads

·  Monthly data ingestion

·  Combining historical + current data

·  Faster than SET (no full rewrite)

Step 1: Sort first

proc sort data=toll_fraud;

by Plaza_ID;

run;

proc print data=toll_fraud;

run;

OUTPUT:

Obs	Plaza_ID	Highway_Name	Region	Vehicles_Per_Day	Avg_Wait_Time	Cashless_Ratio	Collection_Date	Revenue	Efficiency_Score	Month	Next_Month	Days_Since	Utilization	Fraud_Flag
1	TP001	NH44	South	45000	3.5	82	01JAN2026	850000	105E3%	1	01FEB2026	48	High	NO
2	TP001	NH44	South	45000	3.5	82	02JAN2026	875500	105E3%	1	01FEB2026	48	High	NO
3	TP002	NH48	West	38000	4.2	78	01JAN2026	720000	70571%	1	01FEB2026	48	High	NO
4	TP002	NH48	West	38000	4.2	78	02JAN2026	741600	70571%	1	01FEB2026	48	High	NO
5	TP003	NH16	East	51000	2.9	88	01JAN2026	910000	155E3%	1	01FEB2026	48	Very High	NO
6	TP003	NH16	East	51000	2.9	88	02JAN2026	937300	155E3%	1	01FEB2026	48	Very High	NO
7	TP004	NH65	South	29000	6.8	60	01JAN2026	560000	25588%	1	01FEB2026	48	Medium	NO
8	TP004	NH65	South	29000	6.8	60	02JAN2026	576800	25588%	1	01FEB2026	48	Medium	NO
9	TP005	NH19	North	47000	3.1	85	01JAN2026	880000	129E3%	1	01FEB2026	48	High	NO
10	TP005	NH19	North	47000	3.1	85	02JAN2026	906400	129E3%	1	01FEB2026	48	High	NO
11	TP006	NH27	West	26000	7.5	55	01JAN2026	490000	19067%	1	01FEB2026	48	Medium	YES
12	TP006	NH27	West	26000	7.5	55	02JAN2026	504700	19067%	1	01FEB2026	48	Medium	YES
13	TP007	NH30	Central	34000	5.2	70	01JAN2026	650000	45769%	1	01FEB2026	48	Medium	NO
14	TP007	NH30	Central	34000	5.2	70	02JAN2026	669500	45769%	1	01FEB2026	48	Medium	NO
15	TP008	NH66	South	52000	2.7	90	01JAN2026	970000	173E3%	1	01FEB2026	48	Very High	NO
16	TP008	NH66	South	52000	2.7	90	02JAN2026	999100	173E3%	1	01FEB2026	48	Very High	NO
17	TP009	NH52	North	31000	6.1	62	01JAN2026	600000	31508%	1	01FEB2026	48	Medium	NO
18	TP009	NH52	North	31000	6.1	62	02JAN2026	618000	31508%	1	01FEB2026	48	Medium	NO
19	TP010	NH40	East	28000	7.0	58	01JAN2026	540000	23200%	1	01FEB2026	48	Medium	NO
20	TP010	NH40	East	28000	7.0	58	02JAN2026	556200	23200%	1	01FEB2026	48	Medium	NO
21	TP011	NH75	West	36000	4.8	73	01JAN2026	690000	54750%	1	01FEB2026	48	High	NO
22	TP011	NH75	West	36000	4.8	73	02JAN2026	710700	54750%	1	01FEB2026	48	High	NO
23	TP012	NH22	North	24000	8.0	50	01JAN2026	460000	15000%	1	01FEB2026	48	Low	YES
24	TP012	NH22	North	24000	8.0	50	02JAN2026	473800	15000%	1	01FEB2026	48	Low	YES
25	TP013	NH13	South	49000	3.0	86	01JAN2026	920000	14E4%	1	01FEB2026	48	High	NO
26	TP013	NH13	South	49000	3.0	86	02JAN2026	947600	14E4%	1	01FEB2026	48	High	NO

Step 2:Transpose with BY

proc transpose data=toll_fraud out=toll_transposed;

    by Plaza_ID  NotSorted;

    var Revenue Vehicles_Per_Day;

run;

proc print data=toll_transposed;

run;

OUTPUT:

Obs	Plaza_ID	_NAME_	COL1	COL2
1	TP001	Revenue	850000	875500
2	TP001	Vehicles_Per_Day	45000	45000
3	TP002	Revenue	720000	741600
4	TP002	Vehicles_Per_Day	38000	38000
5	TP003	Revenue	910000	937300
6	TP003	Vehicles_Per_Day	51000	51000
7	TP004	Revenue	560000	576800
8	TP004	Vehicles_Per_Day	29000	29000
9	TP005	Revenue	880000	906400
10	TP005	Vehicles_Per_Day	47000	47000
11	TP006	Revenue	490000	504700
12	TP006	Vehicles_Per_Day	26000	26000
13	TP007	Revenue	650000	669500
14	TP007	Vehicles_Per_Day	34000	34000
15	TP008	Revenue	970000	999100
16	TP008	Vehicles_Per_Day	52000	52000
17	TP009	Revenue	600000	618000
18	TP009	Vehicles_Per_Day	31000	31000
19	TP010	Revenue	540000	556200
20	TP010	Vehicles_Per_Day	28000	28000
21	TP011	Revenue	690000	710700
22	TP011	Vehicles_Per_Day	36000	36000
23	TP012	Revenue	460000	473800
24	TP012	Vehicles_Per_Day	24000	24000
25	TP013	Revenue	920000	947600
26	TP013	Vehicles_Per_Day	49000	49000

16. CHARACTER & NUMERIC FUNCTIONS

data toll_text;

    set toll_fraud;

    Highway_Clean = propcase(strip(Highway_Name));

    Region_Upper = upcase(Region);

    Combined = catx('-', Plaza_ID, Region);

run;

proc print data=toll_text;

 var Highway_Name Highway_Clean Region Region_Upper Plaza_ID Combined;

run;

OUTPUT:

Obs	Highway_Name	Highway_Clean	Region	Region_Upper	Plaza_ID	Combined
1	NH44	Nh44	South	SOUTH	TP001	TP001-South
2	NH44	Nh44	South	SOUTH	TP001	TP001-South
3	NH48	Nh48	West	WEST	TP002	TP002-West
4	NH48	Nh48	West	WEST	TP002	TP002-West
5	NH16	Nh16	East	EAST	TP003	TP003-East
6	NH16	Nh16	East	EAST	TP003	TP003-East
7	NH65	Nh65	South	SOUTH	TP004	TP004-South
8	NH65	Nh65	South	SOUTH	TP004	TP004-South
9	NH19	Nh19	North	NORTH	TP005	TP005-North
10	NH19	Nh19	North	NORTH	TP005	TP005-North
11	NH27	Nh27	West	WEST	TP006	TP006-West
12	NH27	Nh27	West	WEST	TP006	TP006-West
13	NH30	Nh30	Central	CENTRAL	TP007	TP007-Central
14	NH30	Nh30	Central	CENTRAL	TP007	TP007-Central
15	NH66	Nh66	South	SOUTH	TP008	TP008-South
16	NH66	Nh66	South	SOUTH	TP008	TP008-South
17	NH52	Nh52	North	NORTH	TP009	TP009-North
18	NH52	Nh52	North	NORTH	TP009	TP009-North
19	NH40	Nh40	East	EAST	TP010	TP010-East
20	NH40	Nh40	East	EAST	TP010	TP010-East
21	NH75	Nh75	West	WEST	TP011	TP011-West
22	NH75	Nh75	West	WEST	TP011	TP011-West
23	NH22	Nh22	North	NORTH	TP012	TP012-North
24	NH22	Nh22	North	NORTH	TP012	TP012-North
25	NH13	Nh13	South	SOUTH	TP013	TP013-South
26	NH13	Nh13	South	SOUTH	TP013	TP013-South

17. PROC DATASETS DELETE

proc datasets lib=work;

    delete toll_raw toll_clean;

quit;

LOG:

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

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

Deleting intermediate datasets:

·  Saves memory

·  Reduces confusion

·  Improves workflow hygiene

18. BUSINESS INSIGHTS

·  High cashless ratio improves efficiency

·  Low revenue + high traffic = fraud risk

·  Waiting time is a critical KPI

·  Digital tolling directly impacts congestion

19. 5 KEY POINTS ABOUT THIS PROJECT

1.This project covers the complete SAS workflow — from raw data creation and validation to cleaning, ransformation, analysis, visualization, and reporting. It demonstrates how real-world toll plaza data is handled in structured enterprise environments.

2. It calculates key performance indicators such as Vehicles_Per_Day, Revenue, Avg_Wait_Time, Cashless_Ratio, and derives an Efficiency_Score (%). This helps evaluate how effectively each toll plaza operates under varying traffic conditions.

3. The project includes rule-based fraud identification (e.g., high traffic but low revenue scenarios). It simulates how analytics teams detect potential revenue leakage or operational manipulation in infrastructure systems.

4. It integrates DATA step, PROC SQL, PROC MEANS, PROC UNIVARIATE, PROC FREQ, PROC CORR, PROC SGPLOT, APPEND, MERGE, TRANSPOSE, and macro programming. This shows strong command over Base SAS and advanced programming concepts.

5. An intentional data-type error is introduced and corrected to demonstrate real-world debugging, validation, and best practices. This strengthens production readiness and audit reliability.

20. CONCLUSION

The Toll Plaza Operations, Revenue, Efficiency, and Fraud Analytics project demonstrates how SAS can be applied to real-world infrastructure and financial monitoring systems. By simulating daily toll plaza data, the project highlights the importance of structured data management, validation, and analytical modeling. It shows how operational metrics such as traffic volume, waiting time, and digital payment adoption directly influence revenue generation and overall efficiency.

The inclusion of fraud detection logic adds practical business value, as revenue leakage and irregular patterns are critical concerns in toll management systems. Through procedures like PROC SQL, PROC MEANS, PROC FREQ, PROC CORR, and visualization tools, the project transforms raw operational data into meaningful insights.

Overall, this project not only strengthens technical SAS skills but also builds business understanding, analytical thinking, and data governance awareness. It is a strong portfolio- demonstrating both programming proficiency and the ability to translate data into strategic decision-making insights.

INTERVIEW QUESTIONS FOR YOU

·  What is the difference between the DATA step and PROC step in SAS, and when would you choose one over the other?

·  What is the difference between SET and MERGE statements in SAS, and how does SAS handle matching observations during a MERGE with BY variables?

·  How do you debug a SAS program if the log shows warnings or errors, and what is the importance of checking the SAS log before validating output?

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

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 TOLL 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 

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

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 SAS analytics reveal which programming languages are popular but surprisingly difficult to master?

2.Is SDTM DM quality control in SAS the key to avoiding last-minute FDA rejection risks?

3.Can SAS analytics reveal which world tourist places truly attract visitors year after year?

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

About Us | Contact | Privacy Policy