Accessibility - Revision history

Zaccuardi: /* Further Reading */ added reference to recent SSTI report

2018-01-17T16:12:46Z

Further Reading: added reference to recent SSTI report

← Older revision		Revision as of 16:12, 17 January 2018
Line 47:		Line 47:
	==Further Reading==		==Further Reading==
	*"The Why and How of Measuring Access to Opportunity: A Guide to Performance Management." Governors’ Institute on Community Design. January 2017. http://www.govinstitute.org/wp-content/uploads/2017/01/how-and-why-of-measuring-access-to-opportunity.pdf		*"The Why and How of Measuring Access to Opportunity: A Guide to Performance Management." Governors’ Institute on Community Design. January 2017. http://www.govinstitute.org/wp-content/uploads/2017/01/how-and-why-of-measuring-access-to-opportunity.pdf
			*"Accessibility in practice: A guide for transportation and land use decision making" Sundquist, Eric, Chris McCahill, and Logan Dredske. State Smart Transportation Initiative. October 2017. https://www.ssti.us/wp/wp-content/uploads/2018/01/Accessibility_Guide_Final.pdf
			.

	==References==		==References==
	<references />		<references />

Leeor at 22:42, 12 May 2017

2017-05-12T22:42:57Z

← Older revision		Revision as of 22:42, 12 May 2017
Line 40:		Line 40:
	Some tools that measure accessibility are discussed below		Some tools that measure accessibility are discussed below

	*[[Sugar]] is a tool offered by [[Citilabs]]. It measures access using decay functions, creating an access score. Sugar incorporates first and last mile travel on a the pedestrian street network in measuring its total travel time. It offers accessibility measures to many types of destinations using destination data from navigation company HERE and allows for custom weighting of destination ~~data~~. Sugar allows for easy layering of demographic data, but does not include demographic based calculations.		*[[Sugar]] is a tool offered by [[Citilabs]]. It measures access using decay functions, creating an access score. Sugar incorporates first and last mile travel on a the pedestrian street network in measuring its total travel time. It offers accessibility measures to many types of destinations using destination data from navigation company HERE and allows for custom weighting of destination types. Sugar allows for easy layering of demographic data, but does not include demographic based calculations.
	[[File:Analysis-spectrogram.png\|thumb\|Accessibility spectrogram from Conveyal Analyst. Shows the increase in accessible opportunities (in this case jobs) as travel time increases. The wider portions reveal system uncertainty or unreliability caused by low service frequency.<ref>Conveyal analysis-ui documentation. Accessed 20 April 2017. http://analysis-ui.readthedocs.io/en/latest/analysis/#spectrogram</ref> ]]		[[File:Analysis-spectrogram.png\|thumb\|Accessibility spectrogram from Conveyal Analyst. Shows the increase in accessible opportunities (in this case jobs) as travel time increases. The wider portions reveal system uncertainty or unreliability caused by low service frequency.<ref>Conveyal analysis-ui documentation. Accessed 20 April 2017. http://analysis-ui.readthedocs.io/en/latest/analysis/#spectrogram</ref> ]]
	*[[Transport Analyst]] offered by [[Conveyal]] is an open source tool. It offers accessibility information for various cut-off points and creates spectrograms that show the change in accessibility as travel time increases, but does not create a related score. Analyst offers accessibility measures for a number of different destination types, acquired from [[OpenStreetMap]], and allows for demographic overlays.		*[[Transport Analyst]] offered by [[Conveyal]] is an open source tool. It offers accessibility information for various cut-off points and creates spectrograms that show the change in accessibility as travel time increases, but does not create a related score. Analyst offers accessibility measures for a number of different destination types, acquired from [[OpenStreetMap]], and allows for demographic overlays.

Leeor at 22:39, 12 May 2017

2017-05-12T22:39:52Z

← Older revision		Revision as of 22:39, 12 May 2017
Line 31:		Line 31:

	==Destinations==		==Destinations==
	Accessibility analysis measures how long it takes to arrive at certain types of destinations. The most common destination ~~that is~~ measured is jobs. Access to jobs is crucial for employers and employees to maintain a robust economy~~, and peak~~ commute hours tend to be the most congested times of the week, so increasing accessibility to jobs via transit is particularly important. Data sources for jobs are also easy to acquire through the publicly available LODES and LEHD data.		Accessibility analysis measures how long it takes to arrive at certain types of destinations. The most common destination measured is jobs. Access to jobs is crucial for employers and employees and to maintain a robust economy. Peak commute hours tend to be the most congested times of the week, so increasing accessibility to jobs via transit is particularly important when seeking solutions to congestion. Data sources for jobs are also easy to acquire through the publicly available LODES and LEHD data.
	Other destinations are also often sought ~~out~~. These can include restaurants, grocery stores, schools, health facilities, parks, etc. Accessibility to these types of amenities has a serious impact on quality of life, and is therefore important in evaluating transit effects on a community. ~~Also, since~~ only 20% of trips are for a home-to-work commute<ref>"Commuting in America 2013: The National Report on Commuting Patterns and Trends." American Association of State Highway and Transportation Officials. http://traveltrends.transportation.org/Pages/default.aspx</ref>, factoring in other trip destinations has a ~~big~~ impact on how much people will use a transit system. There are no government generated sources for these databases. Some tools use open source options, such as [[OpenStreetMap]], while other acquire proprietary databases. A challenge when working with other destinations is how to group them or weight them. Analyzing accessibility to grocery stores might be relatively easy (although even in this example, classifying grocery stores might be difficult). However, analyzing accessibility to amenities broadly raises questions of which amenities count and which amenities are most important.		Other destinations are also often sought. These can include restaurants, grocery stores, schools, health facilities, parks, etc. Accessibility to these types of amenities has a serious impact on quality of life, and is therefore important in evaluating transit effects on a community. Since only 20% of trips are for a home-to-work commute<ref>"Commuting in America 2013: The National Report on Commuting Patterns and Trends." American Association of State Highway and Transportation Officials. http://traveltrends.transportation.org/Pages/default.aspx</ref>, factoring other trip destinations has a significant impact on how much people will use a transit system. There are no government generated sources for these databases. Some tools use open source options, such as [[OpenStreetMap]], while other acquire proprietary databases. A challenge when working with other destinations is how to group them or weight them. Analyzing accessibility to grocery stores might be relatively easy (although even in this example, classifying grocery stores might be difficult). However, analyzing accessibility to amenities broadly raises questions of which amenities count and which amenities are most important.

	==Population Analysis==		==Population Analysis==

Leeor at 22:36, 12 May 2017

2017-05-12T22:36:29Z

← Older revision		Revision as of 22:36, 12 May 2017
Line 28:		Line 28:

	===Using the street network===		===Using the street network===
	Some tools use the street network to estimate travel time between origin or destination and the transit station. This calculation can much more accurately determine the first mile and last mile travel time and therefore ~~yield~~ a more accurate total travel time. It avoids the pitfalls of cutting off people or destinations just outside of a buffer and effectively recognizes relative distances from stations. However, it requires an individual calculation from each origin to each destination, rather than ~~reserving~~ calculations to between stations, hence a much more computationally intensive process. It also requires an accurate street network that includes pedestrian paths.		Some tools use the street network to estimate travel time between origin or destination and the transit station. This calculation can much more accurately determine the first mile and last mile travel time and therefore yields a more accurate total travel time. It avoids the pitfalls of cutting off people or destinations just outside of a buffer and effectively recognizes relative distances from stations. However, it requires an individual calculation from each origin to each destination, rather than preforming calculations only between stations, hence a much more computationally intensive process. It also requires an accurate street network that includes pedestrian paths.

	==Destinations==		==Destinations==

Leeor at 22:21, 12 May 2017

2017-05-12T22:21:50Z

← Older revision		Revision as of 22:21, 12 May 2017
Line 21:		Line 21:

	==First and Last Mile==		==First and Last Mile==
	First and last mile is a common expression referring to how people get to a transit station, and how they get from transit to their destination. These considerations recognize that the time it takes to travel by transit is broken up into four pieces: Travel to the station, waiting at the station, travel on transit, and travel to destination. More complicated trips may include other elements such as waiting for a transfer, or traveling between stations to transfer. Some tools only look at travel time between stations, with the possible inclusion of wait time (usually calculated as half of headway). In that case they use circular buffers to determine which populations and destinations are accessible ~~from the stations~~. Other tools incorporate travel time to and from the stations by using circular buffers and ~~averages~~ or by calculating travel time on the street network.		First and last mile is a common expression referring to how people get to a transit station, and how they get from transit to their destination. These considerations recognize that the time it takes to travel by transit is broken up into four pieces: Travel to the station, waiting at the station, travel on transit, and travel to a destination. More complicated trips may include other elements such as waiting for a transfer, or traveling between stations to transfer. Some tools only look at travel time between stations, with the possible inclusion of wait time (usually calculated as half of [[headway]]). In that case they use circular buffers to determine which populations and destinations are within a certain distance of stations and therefore considered accessible. Other tools incorporate travel time to and from the stations by using circular buffers and calculate an average travel time to the station within the buffer, or by calculating travel time on the street network.

	===Buffers===		===Buffers===

Leeor at 22:18, 12 May 2017

2017-05-12T22:18:35Z

← Older revision		Revision as of 22:18, 12 May 2017
Line 17:		Line 17:
	===Decay===		===Decay===
	[[File:Decay function.png\|thumb\|A decay function showing the change in weighted values for different modes and destination as travel time increases<ref>"Operationalizing Accessibility: Tools and Practices." State Smart Transportation Initiative. 30 March 2017. http://www.ssti.us/Events/operationalizing-accessibility-part-1-tools-and-practices/</ref>]]		[[File:Decay function.png\|thumb\|A decay function showing the change in weighted values for different modes and destination as travel time increases<ref>"Operationalizing Accessibility: Tools and Practices." State Smart Transportation Initiative. 30 March 2017. http://www.ssti.us/Events/operationalizing-accessibility-part-1-tools-and-practices/</ref>]]
	Decay functions offer a more fine-tuned assessment of accessibility by weighting destinations based on how far away they are. A job that is ~~five~~ minutes away contributes more to accessibility than one that is ~~thirty~~ minutes away, and a job that is ~~thirty~~ minutes away contributes about the same as one that is ~~thirty-one~~ minutes away. Decay functions usually have a cut-off, somewhere around the 60 to 90 minute mark, where destinations are considered so inaccessible that they ~~can~~ no longer ~~greatly~~ contribute.		Decay functions offer a more fine-tuned assessment of accessibility by weighting destinations based on how far away they are. A job that is 5 minutes away contributes more to accessibility than one that is 30 minutes away, and a job that is 30 minutes away contributes about the same as one that is 31 minutes away. Decay functions usually have a cut-off, somewhere around the 60 to 90 minute mark, where destinations are considered so inaccessible that they no longer contribute to accessibility at all.
	Using decay is challenging for a number of reasons. A meaningful and reliable decay function has to be developed. The ~~computation~~ effort of calculating the exact ~~distance~~ to each location and weighting the location is ~~large~~ and ~~complicated~~. Finally, because locations are weighted, the final output is unit-less; it cannot be explained as number of jobs within a 30 minute travel time. This may allow for internal comparisons but may be difficult to convey conceptually at public presentations. One solution to improve presentation is ~~developing~~ a clear scale and ~~creating~~ references as explanations for various points on the scale. (ex. a 1-100, where 80-100 is excellent accessibility, etc.)		Using decay is challenging for a number of reasons. A meaningful and reliable decay function has to be developed. The computational effort of calculating the exact travel time to each location and weighting the location is intense and often slow. Finally, because locations are weighted, the final output is unit-less; it cannot be explained as the number of jobs within a 30 minute travel time window. This may allow for internal comparisons but may be difficult to convey conceptually at public presentations. One solution to improve presentation is to develop a clear scale and to create references as explanations for various points on the scale. (ex. a 1-100, where 80-100 is considered "excellent accessibility", etc.)

	==First and Last Mile==		==First and Last Mile==

Leeor at 22:15, 12 May 2017

2017-05-12T22:15:06Z

← Older revision		Revision as of 22:15, 12 May 2017
Line 13:		Line 13:
	===Time cut-off===		===Time cut-off===
	Time cut-offs are useful because they are slightly easier to calculate. Some tool is needed to figure out how far one can go on a system in the specified amount of time, and then all points corresponding to a destination within that area are counted up. They also tend to be easier to communicate to the public or to decision makers. 10,000 jobs accessible in 30 minutes from a given area, or a new line or faster service increasing the number of jobs available in 30 minutes of travel is a relatively easy idea to understand.		Time cut-offs are useful because they are slightly easier to calculate. Some tool is needed to figure out how far one can go on a system in the specified amount of time, and then all points corresponding to a destination within that area are counted up. They also tend to be easier to communicate to the public or to decision makers. 10,000 jobs accessible in 30 minutes from a given area, or a new line or faster service increasing the number of jobs available in 30 minutes of travel is a relatively easy idea to understand.
	Time cut-offs suffer a few disadvantages however. A large employment center just outside of the time range is completely disregarded. So if there are 1,000 jobs available in 30 minutes but 2,000 jobs available in 31 minutes, the analysis will ignore the jobs just outside the cut-off and yield misleading results. Cut-offs also fail to discriminate between where destinations are located within a region. For example, this type of analysis will provide the same results if all of the jobs are five minutes away or thirty minutes way, although being five minutes away clearly provides better access. Some of these disadvantages can be overcome by conducting multiple analyses with different cut-offs, such that instead of having one number for a certain distance, the change as ~~one gets farther away can be observed~~.		Time cut-offs suffer a few disadvantages however. A large employment center just outside of the time range is completely disregarded. So if there are 1,000 jobs available in 30 minutes but 2,000 jobs available in 31 minutes, the analysis will ignore the jobs just outside the cut-off and yield misleading results. Cut-offs also fail to discriminate between where destinations are located within a region. For example, this type of analysis will provide the same results if all of the jobs are five minutes away or thirty minutes way, although being five minutes away clearly provides better access. Some of these disadvantages can be overcome by conducting multiple analyses with different cut-offs, such that instead of having one number for a certain distance, one can observe the change as travel time increases.

	===Decay===		===Decay===

Leeor at 22:12, 12 May 2017

2017-05-12T22:12:52Z

← Older revision		Revision as of 22:12, 12 May 2017
Line 9:		Line 9:

	==Decay vs. Cut-offs==		==Decay vs. Cut-offs==
	One major distinction in measuring accessibility is how parameters are set for determining whether a destination is accessible. One method is to create a time cut-off. For example, one can ask, how many jobs can be reached in 30 minutes using transit from a given location? ~~Or,~~ how many household/job combinations are there within 30 minutes of each other along a transit line or in a transit system? The alternative is to use a decay function. A decay function weights destinations based on how far they are, so that farther locations are given less weight than closer ones.		One major distinction in measuring accessibility is how parameters are set for determining whether a destination is accessible. One method is to create a time cut-off. For example, one can ask "how many jobs can be reached in 30 minutes using transit from a given location?" or "how many household/job combinations are there within 30 minutes of each other along a transit line or within a transit system?" The alternative is to use a decay function. A decay function weights destinations based on how far they are, so that farther locations are given less weight than closer ones.

	===Time cut-off===		===Time cut-off===

Leeor at 22:11, 12 May 2017

2017-05-12T22:11:27Z

← Older revision		Revision as of 22:11, 12 May 2017
Line 9:		Line 9:

	==Decay vs. Cut-offs==		==Decay vs. Cut-offs==
	One major distinction in measuring accessibility is how parameters are set for determining if a destination is accessible. One method is to create a time cut-off. For example, one can ask, how many jobs can be reached in 30 minutes using transit from a given location? Or, how many household/job combinations are there within 30 minutes of each other along a transit line or in a transit system? The alternative is to use a decay function. A decay function weights destinations based on how far they are, so that farther locations are given less weight than closer ones.		One major distinction in measuring accessibility is how parameters are set for determining whether a destination is accessible. One method is to create a time cut-off. For example, one can ask, how many jobs can be reached in 30 minutes using transit from a given location? Or, how many household/job combinations are there within 30 minutes of each other along a transit line or in a transit system? The alternative is to use a decay function. A decay function weights destinations based on how far they are, so that farther locations are given less weight than closer ones.

	===Time cut-off===		===Time cut-off===

Leeor at 22:10, 12 May 2017

2017-05-12T22:10:50Z

← Older revision		Revision as of 22:10, 12 May 2017
Line 6:		Line 6:

	==Accessibility vs. Mobility==		==Accessibility vs. Mobility==
	Transit systems have been increasingly focused on accessibility when evaluating their systems and new projects. Although accessibility has always been a chief goal of transit, past measures have tended to focus on mobility, looking at the number of people who can access the transit system, system capacity, and line speeds. While improvements in service strengthen the transit system, these measures do not evaluate whether people have access to important destinations. Fast transit lines reaching many may nevertheless fail to connect people to job centers, grocery stores, or schools. Far reaching transit systems may force multiple transfers to reach a destination, such that the trip becomes ~~unreasonable~~ long. Accessibility directly addresses these questions by measuring people's ability to reach desired destinations.		Transit systems have been increasingly focused on accessibility when evaluating their systems and new projects. Although accessibility has always been a chief goal of transit, past measures have tended to focus on mobility, looking at the number of people who can access the transit system, system capacity, and line speeds. While improvements in service strengthen the transit system, these measures do not evaluate whether people have access to important destinations. Fast transit lines reaching many may nevertheless fail to connect people to job centers, grocery stores, or schools. Far reaching transit systems may force multiple transfers to reach a destination, such that the trip becomes unreasonably long. Accessibility directly addresses these questions by measuring people's ability to reach desired destinations.

	==Decay vs. Cut-offs==		==Decay vs. Cut-offs==