Arc Collector Part 2: Yellow Light Analysis

Introduction

  The goal of this lab is create, deploy, and analyze data collected using the Collector for ArcGIS app similar to the previous project completed in the Arc Collecor: Part 1, Gathering Weather Data lab. In this project, 24 intersections will be studied around the western portion of Eau Claire to see what the relationship is between road hierarchy, speed limits, and duration of yellow lights. The attributes collected for this project include RdName1, RdName2, Rd1SpdLmt, Rd2SpdLmt, Rd1Type, Rd2Type, Rd1LightLngth, Rd2LightLngth, Time, and Notes. These attribute names represent the name of the two roads at the intersection, the speed limits of the two roads, the hierarchy of the two roads, the yellow light length of the two roads, the time the point was collected at, and any notes. For this project, only intersections with two intersecting roads were studied. Also, only two yellow light times were taken for each intersection. One for each road. It was assumed that the light for the same road in both directions would remain the same.

 The study area is shown below in figure 9.0. The 24 intersections were chosen at random, but a mix between road hierarchies was made sure to have been present for the intent of the study purpose. The study area ranged from downtown Eau Claire to the traffic intersections located along highway 12 (Clairemont Ave). There were many buildings located in the downtown part of the study area compared to the intersections located along and near Clairmont Ave.

Fig 9.0: Study Area Map

    The process of setting up the Collector for ArcGIS app will covered. This consists of creating a geodatabase, domains, setting up the feature class, and publishing it to ArcGIS online. The process of collecting data will be discussed as well. Then, using the attributes collected in the field and making meaningful information from them for the creation of maps will be examined. The results section of this lab will include a series of maps which will be used to help explain the study question and the findings of this project.

Methods

Setting Up the Collector for ArcGIS App
  First, a new file geodatabase was created. Then, the domain properties of the geodatabase were altered to help make the data collection in the field easier. Figure 9.1 shows what the window looks like when creating domains. In total, three domains were created. One for the yellow light length, one for the road type, and one for the road speed limit. The LightLength domain type was range and the datatype was set to float. The allowable range of values was set to numbers between 0 and 10. The RdType domain type was coded values and the data type was set to text. The coded values used include County Road (CtyRd), State Highway (StHwy), Interstate Ramp (IntRamp), Residential Road (ResRd), and Other. The Speed Limit domain type was also set to range values, but this time the data type was set to short integer. The range of values was set from 0 to 100.

Fig 9.1: Geodatabase Domain Properties

  Then, a new feature class called IntersectionInfo was created in the geodatabase. When creating the feature class, many of the properties had to be altered from the default settings. First the coordinate system WGS 1984 Web Mercator (auxilary shpere) had to be used so that the map could be used in the Collector for ArcGIS app. Then, the attribute information had to entered which is pictured below in figure 9.2. The proper data type had to be assigned for each so that the proper domain could be applied correctly. The LightLength domain was applied to the Rd1LightLngth and Rd2LightLngth attributes, the RdType domain was applied to the the Rd1Type, and Rd2Type attributes, and the SpeedLimit domain was applied to the Rd1SpdLmt and Rd2SpdLmt attributes. Applying these domains to the feature class would prove to be helpful when collecting data in the field. Also, by assigning number data types (short integer and float), numeric operations can be performed once data is entered into these fields.

Fig 9.2: Applying Domains to the InterectionInfo Feature Class

  Next, the feature class layer had to be published to ArcGIS online. This was done by first, signing in on ArcMap to ArcGIS online using the enterprise login. Then, by navigating to File → Share As → Service the share as service to ArcGIS online window popped up. This is where the layer settings get set for when it gets published to ArcGIS online. The settings allowed included to create new features, query features, and to update features. An item description was added along with the tags Geog 336, Yellow Lights, and  Intersection Analysis. Then, the publish button was clicked which published the layer to ArcGIS online. The last step to get the feature class ready for editing using the Collector for ArcGIS app was to open and save the map in the map viewer in ArcGIS online.

Collecting Data Using the Collector for ArcGIS App

  Data was then collected using the Collector for ArcGIS app. A bike was used as transportation to travel from intersection to intersection. In total, the data was collected in about 3.5 hours from 11:00 am to 2:30 pm and about 12.5 miles traveled using the bike.

  The yellow light length for the lights were measured using the clock app on the I-Phone 7 using the stop watch feature. To make sure that the yellow light length was accurate, several times were taken per light or until it was felt that an accurate measurement had been made. This was the most time consuming part of the data collection process. The road name attributes were collected by reading the street signs on the traffic lights. The speed limits were found by looking for speed limit signs. If no speed limit sign could be seen from the intersection it was assumed that the speed limit was 30 mph, or if the road had already been used, that it had the same speed limit as before.

  Below, in figures 9.3 and 9.4 is what the map and attribute screen looked like in the Collector for ArcGIS app. The attribute information shown in figure 9.3 is for the point where the arrow is in figure 9.3. 

Fig 9.2: Map Used by the app

Fig 9.4: Attribute Information Screen on the app

Reorganizing Data After Data Collection

  After all of the data was collected for the 24 intersections, it was realized that the intended purpose of the study which was to see how road hierarchy, speed limit, and yellow light duration relate with each other, couldn't be solved with the way the attributes were organized. 

  The first issue with the attributes was that the road names and speed limits were not organized based on road hierarchy. They were organized randomly. These were reorganized by adding new fields names LongestLightTime, ShortestLightTime, HighestSpdLmt, LowestSpdLmt, and RelativeRdHierachy. This had to done because for each point collected, there were two yellow light times, two speed limits, and two road hierarchies. The road hierarchy was changed to more of a Boolean data type. Intersections were reclassified as having equal road hierarchies or not equal road hierarchies. Reclassifying the data into these new attributes fields would help to make creating maps and calculating other statistics easier. 

  Also, a mistake was made when originally creating the time field. The datatype was set to text, but in order to map it, the a new field was created with a short integer data type and the same values from the old time field were used.

  Going back to the study question to see how these attributes relate with each other, 4 calculated fields were created in ArcMap. The first one was the difference between the highest yellow light time and the lowest yellow light time by intersection. The second one was the difference between the speed limits of the two roads at the intersection. The third one was the shorter light time divided by the speed limit of the shorter yellow light. Lastly, the fourth one was the larger light time divided by the speed limit of the longer yellow light.

Results / Discussion

  This first map, shown below in figure 9.5, shows the location of the intersections studied. Most of the lights studied were in downtown, but 7 of them were located along Clairemont Ave. Almost all of the 4 way intersections within the study area were accounted for. 

Fig 9.5: Studied Intersections

  This next map, displayed in figure 9.6, shows the duration of the longer of the two yellow lights at each intersection. The map also has labels for each intersections which indicate whether or not the hierarchy at the road was equal or not.

  Based on this map, there is a relationship between road hierarchy and duration of yellow lights. Many of the roads located in the downtown part of the study area had the same hierarchy and had no difference between the yellow light times for both roads at the intersection. On the flip side, all of the intersections which are classified as having not equal road hierarchy are located along Clairemont Ave or Highway 37 and have a difference between the yellow light times between the two roads at the intersection. Interestingly, there was one intersection located on Water St. and 5th Ave which had an equal road hierarchy, but a difference in the yellow light length. However, this is the only case in which this happens. The map shows a that a difference in road hierarchy correlates with a difference in yellow light times.

Fig 9.6: Difference in Yellow light Times Between Roads by Intersection

  The next map, shown below in figure 9.7, shows the speed limit difference between the two roads at the intersection and compares it to the road hierarchy.
  This map looks fairly similar to the difference in yellow light length map. However, in this map, there were quite a few roads in downtown which had difference in speed limits of 5 mph. This because about half of the roads in downtown had a 25 mph speed limit, and the other half had a 30 mph speed limit.
  Every intersection which was classified as Not Equal except for one had a difference in speed limits of 15 mph. The speed limit of the faster roads (Clairemont Ave and Highway 37) were 45 mph and the speed limits of the slower roads were only 30 mph. There is one outlier within the traffic intersections. It is the traffic intersection of Clairemont Ave and Highway 37. The reason why there is a difference between hierarchies at this light where there is no difference between speed limits is because this is where Highway 37 ends which made for collecting attributes difficult to choose.
  Based off this map, a difference in road hierarchy at an intersection is almost certainly going to mean that there is a difference in speed limits between the roads. 

Fig 9.7: Difference in Speed Limits Between the Two Roads by Intersection

  Next, an interactive map was created in ArcGIS online to attempt to normalize the length of the longer yellow light by the speed limit of the corresponding road. The speed limits of the corresponding roads are labeled for reference and analysis. The values represented by the graduated circles are equal to the length of the longer yellow light at the intersection divided by the corresponding road speed limit. This gives a standardized value which makes the yellow light duration values comparable across intersections no matter the speed limit.
  This value can be seen as a time value for which a driver would have time to react. Interestingly, there is greater time for the driver to react when the speed limit is slower. This occurs in downtown mainly, but is really just where the speed limit is 30 mph or lower. A driver would have less time to react when traveling at higher speeds such as along. This occurs on Clairemont Ave and Highway 37 where the speed limit is 45 mph.
  This map points out that although the length of a yellow light generally increased with speed limit, it does not do so proportionally. Instead, the ratio between light length and speed limit decreases as the speed limit increases.

Interactive Ratio Map Between Yellow Light Duration and Speed Limit of the Higher Hierarchy Road

  A very similar kind of map was created in ArcMap, but with the lower hierarchy roads. This map, below in figure 9.8, shows the yellow light duration to speed limit ratio between the lower hierarchy roads at the intersections. Most of the trends shown in the previous map are presented in this map as well. However, there is one major difference. The difference is that there is a decreased yellow light time to speed limit ratio at the intersections located along Clairmont Ave and Highway 37 even though the speed limits are the same as they are in downtown.
  Perhaps this is because these are the intersections where the road hierarchy is not equal. These shortened ratios indicates that these lights for the lower hierarchy roads located along Clairmont Ave and Highway 37 have a shorter yellow light length than the lights near downtown and Water Street.

Fig 9.8: Lesser Hierarchy Yellow Light Time to Speed Limit Ratio

  This last map displayed below in figure 9.9 shows the time at which the data for each point was collected. The smaller the circle, the earlier that intersection was studied. The larger the circle, the later that intersection was studied. The route taken to collect the data can be seen by this map. The data collector started off in the southeast portion of the map and then collected data up though downtown, crossing the river on Madison street over to Clairmont. From there the data collector went back to collect data for two intersections along 5th Ave and then finished with the final four intersections located along Clairmont in the southern portion of the map.
  It is possible that the duration of yellow lights changes through out the day. There was one instance where the length of the yellow light changed while attempting to measure it several times. If this can happen at one intersection, it could certainly happen at others. For this reason, it was important that the data be collected quickly so that the variable of time didn't become a major contributing factor to the length of yellow lights.

Fig 9.9: Time at Which the Data Were Collected

Other Attribute Information
  The other attribute information such as notes and the street names was either used to help create the new fields to help explain the correlations, or didn't contain enough information where a map could be created.

Conclusion

  The results of this lab found many correlations between the attributes of road hierarchy, road speed limit, and length of yellow lights. It was discovered that when the road hierarchy of two roads at an intersection isn't equal, there is likely to be a difference in the length of the yellow light times. This also indicates that roads with a higher hierarchy have a longer yellow light and that roads with a lesser hierarchy have a shorter yellow light. It was also found that when the speed limit between two roads is different the difference between the yellow light duration is also different. Another important finding was that where the hierarchy of roads isn't equal, the ratio between yellow light time and speed limit was lower than when the hierarchy of roads was equal.
  If this project was to be done over again, a couple adjustments would be made when setting up the attributes. First, the road names would be organized by hierarchy where the road with the higher hierarchy would be always placed in Rd1Name and the road with the lower hierarchy would be placed in the Rd2Name. The second main change would be the road hierarchy classification. Instead of actually listing the road hierarchy, a relative hierarchy rank could be given. These values would be Greater, Equal or Lower.
  If one were to expand on this project. That person should should make the attribute changes listed above. They could also try to collect data on the age of the stoplight. Different aged stoplights are likely to have different yellow light times. Also, the study area could be larger, so that many more intersections could be looked at.