It all started when my wife and I joined a senior center offering free line dancing classes a few months ago. We had met learning ballroom dancing and wanted to give it a try. About 30 seniors were all well engaged and interacting during the 90-minute sessions.

There would be, on average, 4 to 5 different dances to practice in class, and upon coming home, most of the time, we could not remember how the dances went, let alone their names. Like other members, we researched videos on YouTube and tried to practice them. We found it tricky moving back and forth to control the playback and trying the steps within the room, away from the furniture and the computer.

When I learned about this Challenge less than 2 months ago, it totally clicked: Why not use Alexa to control the video precisely to play the exact segments we wanted while trying to learn the steps?

Our instructor, Belle Hodnick, said many members didn't practice much between sessions and made very slow progress. She remembered another instructor who charged $45/mo per member for group lessons and taught the same dance for months to ensure actual progress. Unfortunately, people ended up dropping out.

I surveyed and talked to as many seniors as possible, learning that because it was too hard to practice at home, they would resign themselves to feeling somewhat embarrassed in class but would still want to go for the social aspects and the relatively safe workout. If they could have a magic wand to help them remember well a few dances, they would certainly love it.

Belle and I both felt there was something we could do together to solve this issue by combining our respective experiences.

What It Does

We designed AirDancer primarily to assist seniors who participate in a line dancing class and want to rehearse at home before the next class. We wanted to partner with instructors who would download their own instruction videos for their members to practice at home the dances introduced during class by using our Alexa skill.

Upon starting AirDancer the user is presented with a suggestion to practice one of the dances introduced in class, chosen randomly from a list of 5 dances initially included in the skill.

If the user doesn't like the suggestion, they can decline, and AirDancer will propose other choices in a progressive disclosure fashion, so the user would know how to ask directly for the dance they want next time.

When starting to practice a dance, the user can choose to practice by section, where one section is an 8-count sequence of steps that correspond to eight beats of music. It is a bite-size chunk that is easy to understand and to learn by repeating the section as many times as needed before moving on to the next section to repeat until it becomes familiar.

The detailed instructions are of two types: first, by describing briefly what to do at a slow cadence, and in a second pass, by counting the steps at a slightly faster tempo. If the Alexa device also has a display, it would show the instruction videos synchronized with the audio instructions to make it much easier to understand how to perform the dance.

By repeating selected sections or groups of sections multiple times at will without leaving their position on the practice floor, users can more easily understand what to do at their rhythm. Ultimately, they can ask AirDancer to play a full set from top to bottom to complete the practice.

How We Built It

Our goal from the start was not to try and use Alexa to teach line dancing to seniors in replacement of a live instructor. Our vision is to leverage Alexa's excellent capabilities to augment the benefits of line dancing to all participants, learners, and instructors worldwide. We wanted something that helped seniors enjoy their line dancing classes even more and that instructors would be able to use to track members' progress and even scale up more efficiently to improve their business.

We named our skill AirDancer as its primary benefit is the ability for users to control the delivery of the dance instructions by voice, "over the air," akin to a conversation with a private instructor.

AirDancer's architecture leverages Alexa devices that can be used with a display to provide on-demand fine-grain instructional videos and uses AWS to implement a cloud-based dashboard for instructors to upload their instructional videos to be easily accessible to their members after class. The architecture includes a simple mechanism to bookmark various sections breaking down the dance into easier-to-learn sequences.

We needed to control video playback by voice within the task of learning to dance, and we also spent a lot of time on dialog design to ensure AirDancer would use the proper terms and be easy to use by seniors with varying degrees of computer proficiency.

We are a 3-member team and came together within 6 weeks to build this skill from scratch. We had not worked together at all before we started this project. I met our leading developer, Taka, over the internet, thanks to DevPost. He is based in Japan. Belle Hodnick, now our CDO, Chief Dancing Officer, started teaching line dancing when her instructor moved out of the area, and other members begged her to take over. She is so passionate about line dancing that she decided to do it pro bono, which has been going on for 12 years.

I have a long experience leading small to large software teams (80+) in projects developing seamlessly integrated hardware and software with particular attention to ease of use and user experience for corporate and consumer applications. My wife and I both enjoy learning line dancing and can't wait to use AirDancer to learn more dances.

Challenges We Ran Into

When we formed the team, we had six weeks starting from scratch to submit our skill; we had never worked together before. Thanks to the Internet, and all the wealth of online tools for remote teams, including AWS and Alexa development environment, we were super effective despite working in two time zones separated by 16 hours. Our team members' respective backgrounds and experiences are very complementary, with a deep collective understanding of users' needs and wants, technical architecture, user experience design, subject matter expertise, Alexa skill development, programming talent, and passion for excellence.

We struggled initially with technical and copyright issues as we thought about incorporating music or video content from the internet to augment what users would already naturally do for their usage. Eventually, we built a business model that doesn't require that in its first incarnation. However, with the proper licensing agreements, it still has the potential to include such features when the time comes for an even better user experience and a potentially significant financial contribution to the music and video arts industry.

Of course, the most difficult challenge was ensuring the tool was easy to use and engaging seniors sustainably over time. As a former student of BJ Fogg, the Stanford expert in persuasive technology and behavior change, I was particularly appreciative of all the smart voice interface design guidelines published by the Alexa team. We tried to implement as many as appropriate, considering the short time frame available. Belle and I reviewed the dialogs closely, and we ran a couple of low-tech user testing to ensure we stayed on the right track. We know we could have done even better with more time and resources.

Accomplishments We're Proud of

One critical technical issue we have solved with a simple first-level solution was the need to enable fine voice control of the instruction videos in hands- and feet-free fashion. Our approach could also be adapted for many other instruction videos, including, but not limited to, recipes, knitting, woodworking, repairs, etc....� We intend to use AI in the future to preprocess videos automatically and enable this feature.

We are particularly proud of our ability to work as a team laser-focused on results and of our potential to bring more joy and meaning to the lives of many seniors by developing in record time a fun and valuable skill that has multiple avenues for growth while bringing sustainable benefits to all participants of the line dancing movement. We are very enthusiastic about being part of the Alexa-inspired voice interface movement, poised to be the preferred way to communicate with robots in the future.

Lessons Learned

We started with many ideas for great features, but we quickly realized there was not enough time to implement them. We had planned to connect automatically with the Amazon Music library to enable users to play the right song to practice with music while AirDancer would count the steps to guide them. We couldn't find an appropriate technical solution within the timeframe. We found a low-tech solution by simply suggesting to the user to ask Alexa directly for the right song, independently of AirDancer.

We found the Alexa development environment and the available API very rich. We feel like we have only scratched the surface of what we could do with future versions of AirDancer. One video playback feature we felt was missing from the Video API was the ability to playback at slower speeds (25%, 50%). This missing feature would be a tremendous help for our users.

What's Next For AirDancer

The problem we have set out to solve is not new, and seniors' needs in this space are underserved. Similarly, the business needs of their instructors, who are likely to be seniors themselves, are also unmet by available technologies. We believe we have just started to scratch the surface of this intriguing opportunity.

There are three main growth avenues for AirDancer:

1 - More engagement from seniors - As AirDancer gathers anonymously more usage data from seniors, instructors can leverage gamification features such as streak tracking, proficiency level from the number of practices or number of dances learned, certifications by the instructor. - Seniors voting for their preferred dances will have the opportunity to connect with others of similar preferences, either virtually or in person, through the group classes.

2 - More subscribers from more instructors - As AirDancer proves helpful for seniors to gain more enjoyment and motivation in attending line dancing classes, there will be an opportunity to enroll more instructors to create their instructional videos for their members and leverage other instructors' videos. - As more instructors come on board, they will naturally encourage their students to use AirDancer as a natural companion to their classes. As seniors travel or move to different communities, they will naturally help spread the word to other instructors, thereby feeding the viral spread of the skill. - International markets where Alexa is already present are potential avenues for expansion for AirDancer

3 - Marketplace for instruction content from Instructors - There's an opportunity for an ecosystem where subscribing instructors will benefit from the AirDancer skill to engage their members by improving their motivation for practicing line dancing, thereby growing and retaining their paying audience. - Based on our research and feedback from AirDancer users, we will develop a protocol for building very effective instruction videos that subscribing instructors can use to improve the effectiveness of their classes, thanks to the AirDancer ecosystem. - Instructors who can make excellent content for popular dances will also be able to distribute them through other instructors to more seniors worldwide

Built With

+ 2 more
Share this project:


posted an update

Since our first submission to the Challenge, we have continued to work hard to improve the skill. We are happy to announce that we are about to release a new version with a SloMo option. It is the top most request from our followers, and will help seniors learn a particular tricky dance move when needed. We're wringing out the last bugs and expect to be able to submit the skill for re-certification very soon.

Our next big step afterwards will be to see how we can leverage off Alexa Conversations to enable an even more natural dialog to determine how a user wants to practice, in any way that fits their learning style.

Stay tuned!

Log in or sign up for Devpost to join the conversation.