ICBC Crash Report from 2018 to 2022 

 

Links to Our Project

Interactive Map link:

Final Police Reported Map: Dashboard 1 – Tableau Cloud

Figma link: https://www.figma.com/proto/9jDv4Hv4gtFASMASFQVs3I/INFO-419-Infographic-Final

 

Introduction 

Transportation is an essential part of each person’s daily life. You could be driving, walking, busing or biking. But when we are doing these things we never think of the possible accidents that can occur.  We are all at risk when we are out on the roads. The Vancouver Sun has reported that over the past 5 years there have been more that 400,000 car crashes in the lower mainland (Griffiths, 2021). During the years from 2000 to 2019 Statistics Canada reported that the number of registered vehicles in Canada has increased from 17 million to 25.4 million (Government of Canada, Statistics Canada, 2022). Every day there are over 760 crashes in BC (Injury Topic – Road Safety | BCIRPU, n.d.). It is evident that with the increased cars on the roads this is an important topic to analyze and discuss.  By examining the crash reports provided by ICBC we want to reveal the hazard when it comes to being on the roads and come up with potential solutions. This information would also inform the public of what to be aware of when they are on the road. 

 

Objectives 

The objective of our InfoVis has 2 main goals. The first one is from a high-level action perspective. At this level we want our audience to gain new knowledge from our visual analysis. We want our audience to uncover and be able to comprehend the patterns that they are not able to see in the spreadsheets. We mainly want our audience to discover what areas in the lower mainland have the highest number of car accidents, what weather and road conditions cause the most accidents. By providing this knowledge to the general public about this, it will inform them about the potential risk when they are navigating the roads as well as helping them make a safer choice. The second goal of our InfoVis is at a mid-level action perspective. We want policy makers and the local authorities to look at the visualizations we have made and highlight the importance of it to future drivers. They could implement this into the writing sections of the drivers test.  

Changes

Our work didn’t divert much from our original project plan. Small things came up such as only using the police-reported crashes rather than both police-reported and insurance-reported crashes due to the difficulty in using insurance-reported crash data. At one point we had intended to compare the two data sets to investigate differences in trends, but that idea just wasn’t possible with the datasets, tools, and time we had for this project.

Data Used 

For this project, we used a dataset that was accessed through the DataBC provided by ICBC. This dataset included police-reported crashes from 2018 to 2022. This data is appropriate for our project because ICBC is a credible source and can be referenced. It also operates with the government, and it plays an important role in the insurance industry. They collect a large amount of data in this industry making it reliable and credible. The information that the dataset included collision types, crashes per month and year, crashes in each municipality, what kind of weather the crashes happened in, whether a cyclist and a motorcycle were involved, crash configurations, collision type, if it was a hit and run, if an animal was involved, if a pedestrian was involved, the road conditions, weather, number of casualties, and the total numbers of vehicles involved. We chose to work with their police-reported crashes data rather than all ICBC insurance reported crashes as the dataset was more manageable (smaller in size), cleaner (few to no null values), and had all the data fields we were interested in.

Additionally, we needed to compile a GeoJSON file for the lower mainland of BC spatial data so we could map out the boundaries for our visualization map. While initially we tried using polygon data from the BC government for legal municipality boundaries, not all municipalities written in the police-reports were legal municipalities with clear boundaries. We ended up deriving our own coordinate point data as a CSV spatial file using a mixture of Google maps and latitude.to to identify approximate centers of each area listed as a municipality. While this is an imperfect method of data collection, we went to great lengths to check that each of the map points in google maps, latitude.to, and in Tableau were accurate to the municipality or unincorporated community’s location.

 

Tools Used 

The main tools that we used to design our InfoVis for our project are Tableau Desktop and Figma. In Tableau Desktop we used it specifically to create our graphs. We used Tableau Desktop to create our graphs since it was user friendly to us because we learned how to use it in class and could reach out to our professor for help if we ever run into any difficulties. The other tool we used was Figma.We used Figma to create and infographic. In our original proposal, we were going to use Canva or WordPress but when using WordPress to create a site I ran into some issues with loading the site so I switched to Figma as I had some experience with it and I have never encountered problems regarding the site loading.

As mentioned in the data section, the Government of BC data Catalogue, Google maps, and latitude.to were used for data collection.

Analytic Steps

The data we used required a lot of analysis. We had to create many rough visualizations just to understand the data better ourselves, and ultimately needed many visualizations to communicate the data’s story. First, we picked a couple of things that we thought would be interesting to look at. For example, we knew we wanted to look at which areas had the highest amount of car crashes and how weather affects the amount of car crashes. Then we played with the data in Tableau Desktop to see if there was anything that stood out to us. We included graphs that show what municipalities had the most amount of crashes so people would be aware of this. Then we also included a graph by year and month showing how many car crashes. We wanted to include this because the data we used included the year of 2020, which was during the pandemic. We wanted to see if there was a correlation between the number of crashes and the restrictions of COVID-19. Then we included a graph about how many crashes occur during specific weather. We found this graph very interesting as it shows that there are more crashes on clear days as we thought there would be more crashes on rainy weather as there is an increased chance of occurring. We found that there were more cyclists involved in crashes instead of motorcycles, but that the vast majority of crashes were between vehicles.

 

Design Process 

For the website/infographic we created we first brainstormed a list of interesting correlations in the ICBC data. Then we created a general outline of what we wanted the website to look like.

After we got the basic outline of what we wanted it to look like we then decided on the main colors we were going to use.  We went with the ICBC blue and created a small car logo to match the topic we were talking about.

We then made the visualizations for those data fields while paying attention to the expressiveness and effectiveness of the visualization elements. First we just used the general colors that Tableau Desktop automatically chooses but then we noticed that we should make graphs that have more than one colour, user-friendly to people that are color blind. When using expressiveness and effectiveness in our graph we used color saturation to show which municipalities had the least to most crashes. 

We also used area(2D size) to show the amount of motorcycle vs. cyclist involved.

 

Once we had all of the graphs that we wanted created, we had to think of the flow of the infographic. First, we decided that starting with a small introduction about the topic would be a good way to lead our audience into what would be informing them. Then we dive into the statistics of our graphs, we begin with the trends of crashes over the five years as this sets a general idea of the numbers for crashes and it sets us up to analyze the possible reasons for these numbers we would be discussing later on. Then we would specifically look at municipalities and the number of crashes they had to give us insight into the graphical variation across the lower mainland. Then we would discuss the weather influences and cyclists’ involvement in the crashes and conclude our findings. 

In addition to the infographic demonstrating key information in the ICBC’s police reported crash data, we also wanted to create a more free-form and exploratory visualization of the data. We created an interactive map that allows people to search, browse, and filter through the data field categories to investigate any specific data items of interest to them. Using the map allows people to look at specific areas of the lower mainland that might affect them personally, and see connections between location, crash counts, and any filters they choose to use. While the infographic used a blue theme for colours, orange was chosen for the map to maximize its popout effect from the grey and white background. Especially because most of the data points are of the lightest shade, the orange colour is much more visible than the light blue.

 

The Story & Key Findings 

The two visualizations work together to tell the story of where, when, and how car crashes occur in the lower mainland. The interactive map creates a more accessible way of investigating the data for any trends related to location while the infographic breaks down some of the specific findings that are important to take away from the data set. 

First and foremost, Surrey has a significant amount of police-reported crashes compared to any other municipality. Looking in a slightly larger area, most crashes are in the metropolitan area showing us that population dense areas where there are more cars and tighter roads have more crashes. While this might seem obvious at first, the large difference in the quantity of crashes will hopefully encourage people to drive more cautiously in these metropolitan areas, and for policy-makers to see health and safety value in reducing traffic and cars on the road in these areas.

In our infographic we include a chart depicting a comparison between crashes involving cyclists and motorcycles. We can see that the vast majority of crashes are between two vehicles, but also that cyclists are involved in crashes more frequently than motorcycles. While motorcycles will have to be on the same roads as cars, creating cycling infrastructure to separate cyclists from vehicles could have a very significant impact on reducing crashes. Hopefully this visualization will help people be more aware of the dangers of having cyclists on the road, support such infrastructure change policies, and be more aware of cyclists while they are driving. On the other hand, cyclists viewing this visualization will hopefully understand the significant risk they are putting themselves in when they bike on the roads, and will opt for routes that do have dedicated bike lanes, and to be very cautious at all points where cars could hit them.

 

Pros & Cons

Our visualizations help explore the complicated crash statistics in a more user-friendly, and in a more digestible manner rather than just having to look at an Excel spreadsheet alone. However, one thing that Figma did not allow us to do is embed our map into the site which we did not notice until the very end of completing our project.

Another benefit of having these visualizations is that they are in a much more shareable format than the data sets. By having pictures, figma links, and tableau links, people can share them digitally with ease. The large excel spreadsheets of the ICBC data can be downloaded, but not understood and communicated to people without a fair deal of work. Our visualizations will make the data findings much more accessible, and shareable.

Our visualizations are not perfect. While they do present a certain story and point of view, creating a larger and broader project would provide a lot more relevant context for understanding data points. Car crash numbers can be helpful for understanding the scale of the problem, however measurements of risk, and increases of risk in certain areas would be more accurate for showing the places and conditions that require more caution and policy changes. For example, the data showing weather conditions during these crashes would make more sense within the context of crashes per rainy day, and per clear day. This would require knowing the amount of time during the year it was raining, what time of the day it was raining, and then the time of day for the crashes. With more time, we could’ve gotten a weather conditions data set, analyzed it, and cross-referenced it with the crash statistics to create such an evaluation.

Our map is also limited by the abilities of Tableau interactions. Ideally the map could be more interactive such that zooming in and out of the map would act as a filter for the municipalities included in the column chart, however Tableaus doesn’t have this functionality. Furthermore, With more time we could’ve made a visualization for each data field such that clicking on a filter would add a visualization to the dashboard of crash counts for that data field item. This would increase the interactivity, and the number of conclusions that could be drawn from the map.

Conclusion

This project is fairly to the point. We took ICBC’s police-reported crash data, and visualized it for people to understand the important information hidden away in the large and confusing spreadsheet. Our visualizations make it easy to see that Surrey is currently a high-risk area for driving, and that changes need to be made in policy to target that. You can also see how cyclists are involved in more crashes than motorcycles. Finally, we’ve also made it so that people can investigate the data themselves through an interactive map.

 

References 

Griffiths, N. (2021, December 31). These are the most dangerous locations for traffic collisions in Metro Vancouver. Vancouversun. These are the most dangerous locations for traffic collisions in Metro Vancouver

 Government of Canada, Statistics Canada. (2022, November 17). The Daily — Circumstances surrounding passenger vehicle fatalities in Canada, 2019. The Daily — Circumstances surrounding passenger vehicle fatalities in Canada, 2019 

Injury Topic – Road Safety | BCIRPU. (n.d.). https://injuryresearch.bc.ca/injury-priorities/transport-related-injuries/