Ask HN: Evaluate my app idea

[aspired]

Simple idea; Automagically capture data from LCD of Health/Fitness machine.

Mockup Demo: http://www.vimeo.com/15179861

You just finished your workout on Treadmill or done with your health test on Glucose meter etc. You pull out your Iphone, start TrakBug App, select machine you just used, and then point camera towards LCD of machine. TrakBug reads data from LCD using image recognition. To handle non-LCD cases, Trakbug allows you to enter data manually.

App Goals:

-Automate capture data from variety of health/fitness devices

–Analyze data for trends

-Share data

Questions:

-Would this be a useful app?

-Do you think there is a market for this?

-Who would be most benefited from this app?

-What feature am I missing?

Thanks for the feedback!

[lien]

There's a huge market for this, and devices like this are already on the market based on Wi-Fi, rather than image recognition. I know ICON fitness made an accessory last year that interface to standard fitness equipments, and allows you to share data with your Facebook friends.

this would be a really useful app if it's based on image recognition, but the question here is does it work? to do this, you would have to take an image, send it back to your server for computation, and pass information back to the user. you'd have to periodically feed images to your server - would it be every second? or 5 seconds? it'll be quite a CPU-intensive task, so image recognition may not be the answer if fitness companies out there are already doing it Wi-Fi.

wrt to the use model alone, if you make the user hold the device while they're exercising, no one would do that. it'd be too awkward.

And i wouldn't think about allow users to entering data manually.

[HardyLeung]

I think the OP mentioned that the snapshot is taken at the end of the exercise, basically to capture the statistics.

Even the image recognition part is not that difficult, assuming simple LCD display. The problem is whether they are able to handle the universe of exercise machines (or at least say 95% of exercise machines). One possibility is to ask users to submit image and if their machines are not found, ask for their help to take a few snapshots, etc. Given them a badge so they can brag about it (and even help spread the word).

In addition, I'm worried that the type of data provided is highly inconsistent amount machines. Can the OP easily combine incomplete information from different types of exercise machines?

If somehow the OP can manage this, I think it would be a great app, and yes, people are willing to pay money for this app.

[aspire007]

We got a strong image processing team and have figured out our core image processing algorithm. Our algo should be able to handle 80-90% of LCDs with ease.

If user is using a machine for first time, he has three options, (1) Select one of machines listed in app (2) User creates a machine profile (as shown in video link) (3) Search & download from server where we will continue to add new machine profile & also allow users to post their created machine profile

[aspire007]

We expect user to capture data at end of machine usage when final stats are being displaced. User does not have to hold it during entire exercise. We plan to use image recognition to read LCD, we would be as fast as Red Laser App extract data from image.

[lien]

If other people are able to produce with a graph of real time data with Wi-Fi, while yours is only capable of recording one data point, then it may not make the product indispensible enough for people to use, or pay for it. You are basically competing against all fitness companies out there who already have a large ecosystem of partnerships.

let's take this one question: Nike is a standard app that comes with the iPod/iPhone. what is it about your app that makes people want to use it, as a complete replacement or as an additional value to this app?

i think to approach this one problem, I would probably draw the focus of my app into an "analysis" piece, and steer away from solely being a "data recording" piece. By being an analysis piece of software, you could look into fitness devices out there, pull data from different types of fitness devices: from exercise equipments, to fitness accessories. this could potentially be a huge win if you could pull it off. it doesn't mean that you can't provide image recognition as an additional feature, it just means that's not where the core value of the app is.

Second question, there are already these fitness social networking communities who are already doing some of this already, how do you plan to compete with them? For example, from their front page, dailymile says that it works with garmin, nike...haven't read about what this means, but my guess is that they already have these devices plugged into their social community. what kind of values can you provide here? if your app allows people to record and analyze data no matter what social community they belong to, then it might be pretty valuable.

of course i haven't done much research so i am just pulling things out of my head without giving these things a whole lot of thought. if you start thinking about making a product that is "indispensible" then it will attract a lot of fans

[aspired]

@Lien

ICON fitness & other companies are coming up treadmill, stepper, indoor cycle etc. that can send data over wifi or plugged USB drive or directly to internet. But these machines are expensive & not every professional/personal gym has it.

Completely agree on focusing on "analysis" piece rather than how we get data inside our app.

Comparing with Nike or Runkeeper app or specialized hardware such as Garmin/Sunnto pedometer+watch gets tricky. They target ONLY running. But still let us compare :)

There are dozens of mobile apps for fitness with various feature & price range. (1) Nike app works with external sensor mounted on hardware. Hardware costs $30. Probably runs for 80-100miles, 3-4 months. (2) Some app use on boardGPS some use on board accelerometer such as RunKeeper, but it will not work on treadmill or indoor running (3) App using on board accelerometer (4) Some app allows manual entry of exercise data (5) Garmin/Sunnto professional athlete gear is good in accuracy and shelf life. But it is expensive & cost $120 to $200. Also getting data to PC or online is multi-step process.

Overall, variety of solutions for runners based on their needs. Can our app offer additional value or complete replacement? Nope. BUT are there be ppl who do not want to shell out money to buy Nike/Garmin/Sunto hardware AND use gym machines AND interested to log data, probably yes. How big is the market? I am not sure, and I do not know how to even guesstimate it.

Probably we are better off targeting users of healthcare device such as blood pressure monitor, glucose meter, and such machines. New & slightly expensive machines in market have USB port that allows data downloading to PC. How about offering Smartphone App to allow users to capture data instantly & do all sorts of analytics? No need to plugin your device to PC to download data n analyze. How about we offer our solution to Omron & other device manufacturer for fixed price AND they roll out a free app for their user to instantly capture data giving them competitive edge over other manufacturers in terms of features.

[lien]

icon and fitness manufacturers do make accessories to their standard equipments. also, the thing is that they already have an installed base at gyms. these companies have a large software team and they will be able to eliminate you over time if they decide that this is big enough of a market to enter.

there are industry alliances out there whose single task force is to standardize protocols of how to make it easy for devices to get online. the current model is using the iPhone/iPod as a remote control to get these devices online so it's not that hard. this means even if your app is better, it won't survive for now, but long term it might not if the only true innovation is using image processing to capture data.

Yes, there will be people who will not want to shell out the money to buy these expensive devices. but that means that you are targeting a population who "do not want to shell out the money". You don't really want to build a device where cost is one of your biggest value propositions.

the last idea - I actually really like the idea of targeting healthcare rather than fitness. I actually want to see family members of these patients pay money to track their progress and see what doctors are saying. Doctors are more likely to pay so that they can get access to their patients' data and monitor their patients' progress over time. and yes, partnership is a big thing here because you want to focus on software innovation. by partnering up with hardware companies, you could benefit a lot from them.

there is also a company called Dossia where they keep track of health records. definitely a good partnerhsip to have some time down the road when your product is ready. it's founded by andy grove from Intel, all of my friends who are Intel employees get them for free. I don't know what type of info they keep track here, but if you allow doctors to update info into these health records, then it'd be really useful...now this product sounds really exciting to me! good luck!!

[jonewalker]

I have not done thorough research, but based on what i feel\know, i have following opinion/views.

1. Analytic part : It should not be hard to implement a simple application to generate good figure, pie charts, etc. (much easier than screen/data reading part)

2. Image processing as sole innovation. : Application will solely be based on it as prime innovation/challenge, but that is not so bad thing. Other solutions will be based on wi-fi, blue-tooth, usb, etc and they will have their own limitations (configuring, downloading, connecting etc). Ease of use may decide the winner. Only if all fitness equipment manufactrs. join and develop a universal application targetting large number of machine to connect using blue-tooth, wi-fi etc, then it may be better solution. I donot see this thing happening very soon. (span of 5 years, my estimate). In this time, this application (based on image processing), if it can establish itself, it may possibly be hard to replace it. It would have already created customer base for itself.

3. Application will target low income group. May not be true. If application works as i see it should, i feel it may be used by all groups.