Hello, everyone. Welcome to NEXT ACADEMY Podcast. Today's topic is the making of the independently developed hiking app yamago. Why did Masa, who runs a healthcare business, decide to build a hiking app himself? We will walk through the process from design and testing to release.
Hiking GPS app yamago (free to download | PRO ¥500/month or ¥5,000/year)
https://apps.apple.com/jp/app/id6805208736
Why Build a Hiking App Yourself?
HAL: Let's start at the beginning. What prompted you to make your own hiking app?
Masa: I already went hiking and tried some existing hiking apps. Honestly, though, they had a lot of features I did not need, and many had designs that were not really to my taste. So I thought I would try building something small myself.
HAL: Which features in the existing apps specifically felt unnecessary to you?
Masa: I do not use social media at all, and I have no desire to belong to a community. So the social sharing and showing one another our hiking records were the parts I liked least. On the other hand, I have kept up running, and the app I normally use for that was more focused on the act of running, so I used it a lot.
HAL: What did you like about that running app?
Masa: It supported Apple Watch, and I could do everything on the watch itself. It also told me my pace, which helped when I was aiming for a particular time. I thought those two things might be useful for hiking too.
Designing around Heart Rate
HAL: So that is where the idea of doing everything from the wrist came from. For hiking, you chose to display heart rate instead of running pace. What was your thinking behind that?
Masa: I thought it would make sense to connect it with the healthcare service I run. If you have recorded a hike, it would be good to put that record to some use. I thought visualizing how hiking affects your body could also motivate you to hike, so I connected those things.
HAL: You also added a notification for a sudden rise in heart rate. Was there a situation while hiking when you personally felt you needed that?
Masa: That actually came from a request by an older person. They told me it would help prevent accidents, so I added it. Personally, I have not experienced that situation myself yet.
HAL: In what circumstances did you hear that request from an older person?
Masa: A lot of middle-aged and older people hike, and apparently there are very many cases where physical problems arise along the way and they cannot get back down. If you could know about that in advance, it would make planning much easier. Hearing that led me to add it.
Testing on Real Hikes and Analyzing with AI
HAL: I understand you also tested accuracy on actual hikes. You have figures showing elevation-gain errors of 0.7 to 2.5 percent. How did you carry out those tests?
Masa: I actually hike repeatedly, collect the data, and analyze it together with AI. I repeat that process over and over.
One Year and Four Months of Development
HAL: Let me also ask about development itself. What environment and approach did you use to make the app?
Masa: Development actually took about a year and four months. Building it with the AI available last year would have been out of the question, so I worked on it while waiting for AI itself to improve. Implementing Apple Watch support was particularly difficult. If it only needed to work on the iPhone, it would have been easy, but this part was a real struggle.
HAL: What specifically made the Apple Watch implementation difficult?
Masa: Basically, things would not work. There were repeated cases where data would not transfer from the iPhone to the Watch.
HAL: How did you track down the cause when that happened?
Masa: yamago actually has another sister app, a healthcare app. It also handles large amounts of data, such as blood oxygen and heart rate, so I faced similar difficulties there. But as you keep going, you gradually learn more about all sorts of things. Also, AI is not always right. It often did something wrong or reset midway and rebuilt everything from scratch. I made progress by checking it myself again and again.
HAL: AI resetting midway and rebuilding everything sounds quite troublesome for a developer. Did you find ways to prevent that?
Masa: I made sure to save the history on GitHub once a day. I realized that if something went wrong, I could go back to that point and fix it, so I repeated that over and over.
Building It Is Not the End
HAL: People say the barriers to development have fallen considerably. How did that feel after actually building something?
Masa: It has become easier to build things. But to make something properly usable, you still have to use it yourself and keep improving it. I do not think that has changed. Building it is not the end; turning it into a genuinely usable tool still takes a fair amount of time.
The barrier to building something has certainly fallen. But making it properly usable still means using it yourself and improving it repeatedly. Building it is not the end; that is where it begins. I think the year and four months spent developing yamago were precisely that time.
That is all for today. Please join us again next time.
