Kevin Chang Lead Product Designer

Slow motion video analysis

I led the design for Ubersense, an iOS and Android app that lets athletes record themselves in slow motion and use analysis tools to help them improve.

* Acquired by Hudl in 2014

Product Design iOS & Android Sports Tech

How might we improve the way coaches and athletes use video to analyze their technique?
About Ubersense

Ubersense was a startup that democratized slow-motion video for coaches and athletes around the world. Whereas in the past, coaches had to purchase dedicated slow-motion video cameras to analyze their athletes' technique, Ubersense leveraged the high-frame rate cameras in iPhone and Android phones to allow anyone with a mobile device to analyze their technique the way that pros do.

The iOS and Android apps have been downloaded over 4 million times and used by the US Olympic Bobsled team. Ubersense was acquired by Hudl in 2014.

My Role & Responsibilities

I joined Ubersense (then, a team of 7) in 2013 as a front-end developer and quickly took on the role of leading iOS and Android design. Since Ubersense was a small startup, I got take on significant responsibility and own the design process for many features from end-to-end.

My responsibilities included:

  • User research, customer visits
  • User flows, prototypes, wireframes, UI/UX
  • iOS and Android design
  • Collaborate with engineering team
  • Marketing collateral (branding, website, brochures, presentations)
Understanding our users

Because Ubersense allowed users to analyze their technique in slow-motion, it was especially popular for sports that require high finesse, such as tennis, golf, baseball.

In order to understand users better, I interviewed different types of coaches: I took several lessons from golf coaches to understand how they use video analysis, I reached out to power users, and I met with and observed coaches at training centers.

Select Design Challenges
How might we make it easy to tag and organize videos so users can find them later?

Customer feedback showed that users organized their videos in three primary ways:

  • If they were an athlete, they likely only recorded themselves. Some athletes would want to tag videos by a specific technique. For instance, a golfer might tag their video as "9 Iron" or "Putter".
  • If they coached 1:1 (for instance, a private golf coach), their primary tag might be the athlete in the video.
  • If they coached a team or a class, they may have too many videos and athletes to record to take the time to individually tag each video.

I proposed the solutions:

  • If you start recording a video from an athlete's page, then the athlete tag would be filled in by default. In the cases where a coach has to rapidly change the athlete they are recording, it helps minimize the number of untagged videos there are.
  • If you are recording multiple clips at a time, there is a high chance that all of the first frames looks very similar. When you are tagging, you should be able to delete and preview videos that you want to keep.
  • If you are recording multiple clips at a time, there should be a smart default that associates tags to all the clips in a playlist.

How might we improve the feedback that coaches give athletes?

During a lesson, a coach may want use video to show an athlete where they should specifically improve. The coach may verbally communicate these points during a lesson but may want to follow up afterwards so that the athlete can remember them better.

There were several improvements made to enhance communication:

  • Allow coaches to record voiceovers with annotations and share those videos with their athletes. This way, the coach can reiterate the improvement points and the athlete can keep those with them the next time they are practicing.
  • Allow coaches to comment on videos. Sometimes athletes would send coaches videos asking for insights. Instead of having to email back a response, coaches (and anyone else an athlete might send the video to) are able to leave comments so that they are centralized in one place.
  • When recording a voiceover, coaches are able to compare their athlete's form to a professional. Professional videos were made available across a number of sports, including golf, tennis, and baseball.

How might we allow athletes to tap into the larger Ubersense Community?

Customer feedback showed that users organized their videos in two ways:

  • If they were an athlete in a group class, the coach could set up a team space where athletes could share and comment on each others' videos.
  • Users could upload their videos to the broader, public Ubersense Community where they could solicit feedback from others.

How might we bring the mobile slow-motion scrubber to the web?

This prototype was created to test the experience of using the slow-motion scrubber on the web. The key challenge I found was that at that time, iOS videos had a consistent frame rate, but since there are so many different types of Android devices, the frame rate that cameras could capture was too varied. Because the experience couldn't be replicated at a high-enough standard, this prototype showed that it was to early to allow users to scrub videos consistently on the web.

App Architecture

Since Ubersense was built iteratively, it took a while for us to nail down the information architecture. We revised the hierarchy a number of times to ensure that the most important features were accessible to most use cases.


In 2013, the US Bobsled team used Ubersense to train for the 2014 Sochi Winter Olympics. The Ubersene team visited the US Bobsled team at their training center in Lake Placid, N.Y.

In 2014, Ubersense was acquired by Hudl.

© Kevin Chang, All Rights Reserved