Final Report - Madruga, Huber, Guererro

Improving Google Maps Walking Directions

Authors

Kyle Madruga, Issac Guerrero, Tom Huber

Abstract

The goal of our project is to improve Google Maps Walking Directions by allowing users to draw in their own walking paths. Currently, Google Maps cannot guarantee that roads have sidewalks, and cannot give directions which include walking paths and bike paths that it does not know about. Also, Google Maps cannot give directions that include routes that cut through parks, public areas, parking lots and other areas which pedestrians normally use.

We would like to create a system which would allow a user to draw in a walking path, or a park, or indicate that a particular street does not have a sidewalk so that Google Maps can later use that input to give users more reliable, safer, and quicker routes to get where they are going by foot.

Keywords

Google Maps, Walking Directions, Users, Creations, Crowdsourcing

Statement of the Problem

The problem we are looking to address deals with the inaccuracies in the walking directions given by Google Maps. Currently,Google Maps does not take advantage of walking paths, bike paths, and other walkable areas. We are attempting to design a user friendly system which allows users to draw in and label paths which the path finding algorithm used by Google Maps can include when giving future directions. Although Google is continually improving their path finding algorithm, there are simply too many walkable areas (such as parking lots, parks, open lots, etc.) for one company to map. However, it is entirely possible for a large user base to map a significant number of these areas accurately and relatively quickly.

As we have studied this problem, we have found that a significant number of internet users have used Google Maps, and furthermore that most of them trusted the directions given by Google Maps, unaware that there were paths that the mapping algorithm wasn't using. Our original approach was focused on improving the algorithm already in place, with the ultimate goal of having the interface of Google Maps appear the same as it does now, showing only a single path. This notion changed, however, as we interviewed users and received surveys, it seems that a significant portion of our user base is interested in seeing multiple paths which they can decide between. This was unexpected and changed our approach significantly including our interface design, and approach.

Rationale

Through our research we have found that a large percentage of people would be interested in this system. Although users can already use the satellite view to add walkable paths to their directions, people who don't know the area they are walking might end up taking paths much longer than they need to be. Anyone who already uses Google Maps for walking directions could see an improvement in the accuracy of the directions they are given.

Relationship to the Course

Our system is a digital and social system in itself. In order to be successful, and useful, community participation is integral. Our system models a Web Bush in that it provides information to users about various streets and paths, including whether a road is under construction, weather related concerns such as flooding and lightning conditions, and the existence of sidewalks. Furthermore, our system can potentially play the role of the metaphorical calculator which teaches you math, in that our system can inspire intrepid users to explore the world around them, looking for walking paths and parks that can be put into the system, thus improving their overall navigational skills. This makes our system a path finding algorithm that increases your ability navigate the world around you, rather than hindering it. Our system also encourages a sense of community, in that users are improving the Google Maps experience for other users. We have found, by talking with our user base, that our system might also inspire friendly competition among users, where users would gain a feeling of social capital from adding more paths to the system and thus compete with each other to better the community as a whole.

Contribution of Individual Team Members

Throughout the development process, we worked as a team, electing to do work together, rather than separately. Consequently, no single body of work can be credited to only one person.

Description of Independent Research

Google Maps has recently added walking directions, this feature ignores any restrictions that a pedestrian would not need to worry about, but is lacking in information. Google has stated "We're working on collecting new data on pedestrian pathways and on more effective ways to solicit your feedback, so that we can steadily improve this feature and get you where you need to be as efficiently as possible." We believe that our project is an efficient and creative means to this end. Our system would provide a way for users to input and rate paths which are missing from the Google data base, thus improving the path finding system dramatically with very little cost to Google.

Our group has conducted research which has revealed that the Google Maps user base wants the ability to make such contributions, and would be willing to do so. We have also researched how to effectively present our system to potential users so that users feel comfortable with the interface and, at the same time, have the power to drastically improve the Google Maps community.

Our group split this project into three sections: Consuming, Creating and Rating.

Consuming:

Our research related to consumption was focused on what users want from the system. That is to say, what features should our system offer a user that would improve their overall experience and provide the quickest, safest, and most direct routes to where they are going. We decided early on that the kind of paths offered by our system would be important. The current Google Maps interface offers a single route that can be edited by users in a limited capacity, users can say that they want to take a particular road, or avoid one area by placing path makers, but users cannot travel a path which the system does not know about. For instance, a user cannot cut through a park, the path generated would go around the park instead, making distance and time estimations worthless. We found that users would feel more comfortable with a system that shows them several paths, one which is the most trusted path, comprised partially of non-user generated roads, and several paths which include user generated content, giving them the ability to decide for themselves what walkways to use. Users are given sense of empowerment from such decisions, making them feel better about their overall Google Maps experience.

Creation:

Our research regarding creation was focused on what users would find helpful in adding paths to the system. In order to do this, we considered the kinds of things that Google Maps does not know about, and what kind of tools we would need to add these things. We decided that a pen tool would be very helpful and intuitive when creating roads and paths. The next thing that we had to consider was parks and open spaces. For these kinds of areas, a pen tool would not have the power necessary to convey all the possible paths through the area. So, we decided that a polygon tool would be helpful. Users would select the corners of an area and the mapping algorithm would treat that area simply as a straight line path through the area in the necessary direction. We felt that it was important to keep the interface simple and the number of tools to a minimum, while still giving users a large amount of freedom. We considered having highly specialized tools for many situations but felt that this would scare off novice users. Instead, we condensed the number of tools into the two described above. To differentiate types of paths, after a user has drawn in a path, the user would be prompted to add a name, a description, and hazard information to their path which would be available to consumer-users later on. As this becomes more developed, necessary tools would reveal themselves. Since we do not have the ability to implement this system, we do not have a large enough user base to discover such tools, but we feel that if this system were to be implemented, the potential for the offered tools would reveal itself naturally.

Rating:

The final segment of our research was focused on how users can rate paths contributed by other users. This section makes the community that we hoped to create real. We decided that allowing users to rate based on a scale would not be helpful, since each person can rate a path positively or negatively for different reasons, and might weight certain aspects more heavily than others. Instead, we decided to give users a choice between Accurate and Inaccurate, and allow them to add in temporary flags which could contain information concerning reasons that the path shouldn't be used that don't relate to accuracy, such as construction work. Users would also be able to flag a path as inappropriate. This would allow users to help eliminate paths that should not be used, such as ones that don't exist, cut through private property, or uses unsafe street crossings. We also thought of sending out a after-travel survey to users who are logged into a Google Account. We made it clear to our user base that this email would only be sent to users who were already logged into a Google Account, would ask for no information from users and could be disabled at any time. This system would inspire more people to use the walking path system and create an awareness in the community. Furthermore, it would improve the accuracy of results given by the system and improve the overall system. This idea was warmly received by our user base which showed an overall interest in helping the community.

Future Work

If we were given another semester, we would explore the possibility of developing this system in conjunction with Google. Since we do not have access to the data base which Google uses for their path finding algorithm, we cannot create a prototype system which would return any useful results. We would also create a mock-up of the user interface and show it to a varied user base to gauge its effectiveness with regards user experience and technical ability.

User Quotes indirectly referenced in this article

User Survey