Upload
manuel-martin
View
249
Download
0
Embed Size (px)
Citation preview
Improving transport timetables usability for mobile devices:
a case study
Manuel Martin Salvador, Marcin Budka, Tom Quay, Anthony Carver-Smith
11th International Conference on the Practice and Theory of Automated Timetabling24/08/2016 - Udine, Italy
Mobile vs Desktop traffic
Mobile
TabletDesktop
August 2016
TransXChange: UK nationwide standard for exchanging bus schedules and related data.
● bus schedules including stops, routes, departures times, departure frequencies, etc.
● the days on which the services run● bus operators information● information about accessibility of
stops and services for wheelchair and other users
● 316 pages PDF!
Transport data in the UK
NCT bus information system
TransXChange file (XML) Importing script Journeys
DatabaseTimetable generator
Journey planner
Departure boards
Real-time information
Fares UI
UI
UI
NCT bus information system
TransXChange file (XML) Importing script Journeys
DatabaseTimetable generator
Journey planner
Departure boards
Real-time information
Fares UI
UI
UI
ChallengesTimetable generation
- Support TransXChange- Grouping lines with similar routes- Day shifting (buses running after 00:00)- On the move → Dynamic timetables
based on location and time
Timetable UX
- Small screen size (too large timetables) → Compress rows and columns
- Variable resolution (vertical vs horizontal orientation) → Responsive
- User interaction → UI controls
Usability study
● Set some tasks for the user to solve
● Ask the user to think out loud
● Let the user freely interact with the app
● Take notes!
● Optional: eye tracking
Continuous usability testing
Hypothesis Design Develop Test Decide
Continuous usability testing
Hypothesis Design Develop Test Decide
Formulate the initial hypothesis. For example:
“90% of users can find the time to catch the next 1A bus at County Hall stop in less than 20 seconds”
Continuous usability testing
Hypothesis Design Develop Test Decide
Design one or more solutions that help to test the hypothesis. For example:1. Plain timetable2. Plain timetable + search box3. Compacted timetable
Continuous usability testing
Hypothesis Design Develop Test Decide
Implement the solutions in a testing environment.
Continuous usability testing
Hypothesis Design Develop Test Decide
Perform the usability test by selecting a number of users and splitting them in different groups (one group per solution). We should collect as much qualitative and quantitative feedback as possible that can help make a final decision.
Continuous usability testing
Hypothesis Design Develop Test Decide
Analyse the results, present a report to the stakeholders, and decide the next steps.
Continuous usability testing
Hypothesis Design Develop Test Decide
Ideally, the final decision would be to release the best solution to the production environment. However, the decision could be to reformulate the hypothesis, test different solutions, or increase the number of users, among others.
Key performance indicators
American Customer Satisfaction Index (ACSI) indicators:
● Overall satisfaction. How did you feel overall with finding the information in the timetable? From 1 (very dissatisfied) to 10 (very satisfied).
● Level of difficulty. How difficult was it to find the information that you needed? From 1 (extremely difficult) to 10 (extremely easy).
● Expectancy. Do you think that this timetable is the most appropriate tool for finding the bus information? From 1 (falls short of your expectations) to 10 (exceeds your expectations).
Additional feedbackQuantitative. For example:
● Number of taps● Time per screen● Total time
Qualitative. For example:
● User direct feedback (spontaneous or using a questionnaire)
● User indirect feedback (when thinking out loud)
Timetable UX changes
Time navigation buttons
Tap on table to expand/collapse stops column
Header
3 rows direction/date
Button to show all stops
Row and column highlighting
Only main stops are initially displayed
Compact times by frequency
Results3 Tasks: easy, medium, difficult.
Total participants: 12
Splitted in 2 groups:
● Group A: original timetables● Group B: new timetables
3 Key performance indicators:
● Satisfaction● Difficulty● Expectancy
Lessons learntInitial assumptions are often wrong → Test more, test often
Finding people to do testing (for free) is difficult → Incentives could help
There is no “average” user → Find people from different backgrounds
Future workTake advantage of other information sources:
● User journeys (e.g. automatic highlighting based on user common routes)
● Realtime information (e.g. show where are the current buses)
● Disruptions (e.g. adjust times based on planned disruption information)
Accessibility testing with visually impaired users