Introduction and Research Question

Dataset: City of Melbourne on-street parking sensor data, 2019.

Dataset size: 42,672,743 parking records.

Recorded violations: 1,558,893 records flagged as violations, approximately 3.65% of all records.

Analysis tools: R, DuckDB and Plotly.

Research question: How do recorded on-street parking violation rates vary across parking areas, months and hours of arrival in Melbourne during 2019?

This report examines parking sensor records across geographic areas, months and parking arrival hours.

Parking Area Analysis

##      AreaName total_records violation_records violation_rate
## 1   Docklands       5057409            256416           5.07
## 2 Queensberry       3470863            106165           3.06
## 3   Southbank       2691230             91950           3.42
## 4    Jolimont       2415712             74154           3.07
## 5         All       2042347            178257           8.73
## 6      Titles       1910619             67635           3.54
## [1] 38
## Warning: Specifying width/height in layout() is now deprecated.
## Please specify in ggplotly() or plot_ly()

Interpretation: Parking Violation Rates by Area

Key findings to discuss:

  • However, Kensington’s rate of violation is highest at 26.46%, and this is only on 427 parking tickets.
  • A Domain violation rate is 10.40%, with 112,274 records recorded.
  • Based on over 5 million recorded docklands violations, it has a recorded violation rate of 5.07%.
  • Violation rates vary between Melbourne’s parking areas.

Interpretation to develop in your own words:

Describe the difference in the number of parking violations in different areas. Explain the need to be cautious with such a low level of records in Kensington as a result of the high percentage. The Docklands has a significantly higher number of parking transactions.

Limitation: The graph shows the difference in the rates of violations recorded, but it does not indicate what might have resulted in the difference.

##          AreaName total_records violation_records violation_rate
## 1      Kensington           427               113          26.46
## 2          Domain        112274             11675          10.40
## 3             All       2042347            178257           8.73
## 4        Courtney        845717             52009           6.15
## 5          Family        667358             36648           5.49
## 6        McKillop        673030             36599           5.44
## 7       Docklands       5057409            256416           5.07
## 8         Windsor        737352             31702           4.30
## 9        Hardware       1092459             46823           4.29
## 10 East Melbourne        725259             30452           4.20
## 11           RACV        191850              7916           4.13
## 12 West Melbourne        767139             31344           4.09
## 13    Twin Towers        726690             28495           3.92
## 14    City Square        373315             14569           3.90
## 15      Chinatown        542385             19986           3.68
## [1] 12

Interpretation: Monthly Parking Violation Rates

Key findings

  • The rate for February was the highest at 4.30%.
  • April had the least amount of violations in terms of percentage at 3.43%.
  • The highest rate was 0.87 percentage points higher than the lowest.
  • Monthly rates were between 3.44% to 3.68% from May to December. Points to explain in my own interpretation

There are relatively small changes in recorded violation rates throughout 2019 as depicted in the monthly data.

Talk about the pattern suggesting normal rate of violations throughout the year instead of a strong seasonal pattern.

Describe that these differences are not an explanation for why recorded violations increased or decreased.

Limitation: This chart does not reflect differences in parking restrictions, locations or enforcement conditions from month to month.

## [1] "hour"              "total_records"     "violation_records"
## [4] "violation_rate"
## [1] 24

Interpretation: Hourly Parking Violation Rates

Key findings

  • Parking arrivals at 7 AM had the highest violation rate of 7.17%.
  • The recorded violation rate was 6.33% at 10 AM.
  • The recorded violation rate was 1.51% at 6 PM.
  • The violation rate at midnight was a low 0.21%.

Points to discuss in my interpretation

  • As indicated in the results, there are significant differences among the parking violation rates as recorded for the different hours of parking arrival.
  • Compare rates of arrivals at different times of day - rates for morning arrivals and rates for evening / overnight arrivals.
  • Discuss why these patterns might be relevant to the study of parking management, but do not make any unsupported conclusions about the cause of these patterns.

Limitation

  • The hourly analysis groups records by the time the car was parked. Does not set the exact time of a parking violation or the reason for the flag.
## Area records: 42672743
## Monthly records: 42672743
## Hourly records: 42672743
## Area violations: 1558893
## Monthly violations: 1558893
## Hourly violations: 1558893

Conclusion

Summary of Findings

  • There were geographical patterns in the data for parking violations: recorded parking violation rates were different between each parking area in Melbourne. Kensington’s rate was 26.46% based on 427 records, whereas Docklands’ rate was 5.07% based on over five million records.

  • Monthly violation rate varied, with February being the highest at 4.30% and April the lowest at 3.43%. The difference was 0.87 percentage points.

  • Hourly patterns: The most prolific parking violation occurred with parking at 7 AM, which had a violation rate of 7.17%. Evening and overnight arrivals, overall, generally had lower rates.

Overall Interpretation

Geographical variations: Violation rates are different across Melbourne’s parking areas. Kensington has the highest observed rate of 26.46%, however, there are only 427 records available for this statistical group. Docklands on the other hand has a rate of 5.07% with over five million records. This reinforces the fact that the number of parking tickets issued at a location and violence rates are significant factors to consider when comparing locations.

Monthly trends: The numbers of violations have been relatively steady throughout 2019. February is the peak month at 4.30% and the lower being at April with 3.43%. The variation of 0.87 percentage points indicates that month-to-month changes are fairly small.

Patterns of hourly arrivals: The pattern of arrival time is more clearly seen as hourly patterns. The average violation is highest for the Parking Record at 7 AM at 7.17%, and lowest at midnight at 0.21%, and at 6 PM at 1.51%. These numbers are for times of parking, however and not times of the violation.

The big picture: Of the 42,672,743 parking records you’ve analyzed, including 1,558,893 records marked as violations, there are noticeable variation in parking violation rates based on location and arrival time, and monthly rate variation is not as significant. These results outline trends in sensor data, but do not determine the causes of the trends.

Limitations

Recommendations for Further Research

Suggest what additional data could help investigate the observed patterns, such as parking restrictions, enforcement information or data from other years.

References

City of Melbourne. (n.d.). On-street car parking sensor data – 2019 https://data.melbourne.vic.gov.au/api/v2/catalog/datasets/on-street-car-parking-sensor-data-2019/exports/csv?delimiter=%2C . City of Melbourne Open Data. https://discover.data.vic.gov.au/dataset/on-street-car-parking-sensor-data-2019