01. Problem Identification

Commuters in the Denver Metro area who have alternating schedules may not always remember the odd train schedules. How do they currently find the times? How might an Alexa voice skill be used to deliver the correct train schedule to someone on a varying day-to-day basis?

Demo Video

goals

Project

Identify a problem that can be solved with the use of a voice interface skill.

User

Did the user get the scheduled time they needed?

Personal

Learn to read and write code specific to voice skill development.

Having attainable goals helped keep me on track for the progression of this project and measure its success.

02. RESEARCH + SYNTHESIS

This project heavily relied on research on how to best develop a voice skill to yield the most preferred route for the user to take. For a proof of concept, I focused on Auraria Campus and surrounding Denver area. This helped focus on a specific type of user.

NEED-FINDING

A need-finding exercise determined frustrations with finding the light rail schedules for commuters and revealed they want an all-in-one stop to find all of the information regarding their total commute.

Frustrations

Lack of accessibility, distant locations of stations and varying schedules while trains were often not on time.

Methods

01. Show up to the station
02. RTD Website
03. Google Maps
04. Apple Maps
05. RTD Trip Planner

vertical comparisons

Uber

Has the most diversity in commands.

Sample Utterance: “Alexa, ask Uber to call me an Uber X from work.”

Travel Time

Works with saying station names well for customization.

Sample Utterances: “Alexa, open Travel Time”; “New York”

New York Subway & Transit Status

Works with real time data and status updates incorporated in the user’s flash briefing. Gives “read more” on card in app.

Sample Utterance: “Alexa, what’s my FlashBriefing?”

Universal Parks & Resorts

Works well with a variety of information from show times to park hours to delays and directions.

Sample Utterance: “What are the showtimes for WaterWorld?

survey

I distributed a survey that consisted of 11 questions, some serious and not so serious. The questions were about personality, day to day life, voice interaction experience, and transportation methods and experiences.

Survey

Top Insight

26/36 people spend over 30 hours away from home during a week.

linguistic research

While it may not seem like Colorado has any known accents, residents do have interesting characteristics, like how to pronounce “mountains.” Unlike some of the other states, such as Texas and New Jersey, one can find distinct pronunciations of letters. The Denver RTD skill is for locals where the main demographic most likely will not have thick accents. There is greater diversity with the rise of people moving into the Denver area; therefore, designing the skill around different dialect variations would be a smart choice.

03. product design

Product Pitch

“Denver RTD” provides the Denver Metro train schedule to your Alexa device. Get updates on when your train is scheduled to arrive at a designated station. You can even add your favorite station. Let us help you get to your destination on time.

PRESS RELEASE

Amazon Web Services paired with the open source data provided by Regional Transportation District (RTD) supplies the resources to build and develop a skill that will help all Denver commuters who have altering schedules, thus creating an efficient solution to keeping up with their commute. Developing thisAlexa Skill provides users full access to the light rail times hands-free rather than scrolling through the browser on their phone or computer every time you need to catch a train.

Full Press Release

Flowchart

This flow had all of the questions being asked by Alexa and the user saying limited answers as well as ‘yes’ and ‘no.’ This was built with the decision tree style. This also established the parameters the user has to go through in order to get all of the information for the Alexa to give an accurate result.

Click on the images to view larger.

Information architecture

Configured for Spring 2018 RTD stations.

Click on the image to view larger.

04. USABILITY TESTING

I utilized screening questions, questionnaire post usability test and also received open-ended feedback during the testing phase. In the usability tests, I measured the success the user had in getting the information they needed regarding what train to catch. I asked how many parameters the user heard and remembered as well as what parameters were important and which ones weren’t. I was given many different answers as to how customized the experience should be. I iterated on the flow of the skill the most based on these tests but also altered the content.

User Videos

Results

Quantitative Data

7/7 users were able to complete the skill from start to finish.
7/7 users were able to start the skill using “Denver RTD”
2/7 users said “please” at some point during the interaction to show their manners.
3/7 users mentioned that they wanted the skill to include the full commute time from their house to their end location.
2/7 users have an Amazon Alexa in their personal homes.
2/7 users have Google Home in their personal homes.
4/7 users have Apple’s Siri.

Surprises

The sassy attitude.
How much information it gave (parameters) even though it didn’t seem like a lot.

Problems

Alexa couldn’t pronounce“a.m.” correctly.
Wanted to be able to customize experience for different scenarios.
Didn’t catch all of the dialogue because Alexa spoke too fast.

Denver RTD

A voice interface skill developed for Amazon Alexa Echo Dot.

TASK

Develop and perform user testing for a voice interface skill for the Amazon Alexa Echo Dot.

Project overview

A voice skill where Denver, Colorado's transit schedule is integrated into Amazon's Alexa to enhance the experience commuters have when their schedule is not routine.

Deliverables

A functioning Alexa skill, final demo video and a case study.