Is it crazy to start a business enabled by raspberry PI

[zcapozzi]

I have an idea that uses PIs and PiCams, but I also have a tendency to latch on to the first technology that I come across instead of finding the best solution. Is it crazy to try and create a business that sells services that are enabled by computer vision + Raspberry Pi?

[piglet]

You'd need to be a little more specific - but if you can get the numbers to add up on the business plan and you have the expertise to deliver...why not?

[zcapozzi]

You are right. From a business plan perspective, it's less about the technology choice and more about the fundamentals.

I guess my question was more about the choice of Pi + Pi Cam as the technology foundation. If the general process is as follows, are there alternative technology platforms I should be evaluating:

1. Capture video
2. Parse video to images
3. Interpret images
4. Process data into results
5. Continue indefinitely

[expandables]

You are right. From a business plan perspective, it's less about the technology choice and more about the fundamentals.

From a business plan perspective, it's all about the profits.

[mikronauts]

If I were you, I'd make a proof of concept device based on a Pi 2 to see if your idea is feasible - ie can the Pi2 meet your requirements for image size, quality and recognition at a high enough rate to do whatever it is you want to do.

There is absolutely no way anyone can tell you if your project is even potentially feasible without a lot more information from you.

Your best bet is to hire a consultant to do develop a requirements document, and do a feasibility study.

Based on past experience, you will likely get sticker shock once you see how much it would cost to do the requirements & feasibility study, never mind the full project.

You are right. From a business plan perspective, it's less about the technology choice and more about the fundamentals.

I guess my question was more about the choice of Pi + Pi Cam as the technology foundation. If the general process is as follows, are there alternative technology platforms I should be evaluating:

1. Capture video
2. Parse video to images
3. Interpret images
4. Process data into results
5. Continue indefinitely

[alexeames]

Your best bet is to hire a consultant to do develop a requirements document, and do a feasibility study.

Based on past experience, you will likely get sticker shock once you see how much it would cost to do the requirements & feasibility study, never mind the full project.

Way to kill a project before it starts.

I'd say forget that approach and make one. If you can get it working right, show it to a few people and see what they think. Then tell them how much you're planning to charge for it and see what the reaction is.

[expandables]

I had a course on entrepreneurship and the lecturer was mainly talking about net-loss and net-income. The question you should be asking yourself is this, can you start a profitable business with a product base on the Raspberry PI.

[mikronauts]

LOL

In some ways I agree with you - but the OP was specifically wanting to start a business, and sounded more like a marketing guy than a developer.

The odds of a marketing guy / MBA type getting a new Pi OS running is so close to zero as to not matter.

Us (as my wife calls me) "techy-geeks" could do it, if it was an itch we wanted to scratch, and we had the spare time...

Therefore, if he did a marketing study, and there was a market, he needs to hire people to make it ... unless he has the time to learn enough to write it, which is probably a much bigger chunk of time than he has available within the perceived marketing window.

Your best bet is to hire a consultant to do develop a requirements document, and do a feasibility study.

Based on past experience, you will likely get sticker shock once you see how much it would cost to do the requirements & feasibility study, never mind the full project.

Way to kill a project before it starts.

I'd say forget that approach and make one. If you can get it working right, show it to a few people and see what they think. Then tell them how much you're planning to charge for it and see what the reaction is.

[zcapozzi]

Sorry, should have been more clear. Prototype is mostly functional (almost). I want to get the prototype working before seriously thinking about business case. To be honest, this is more of a fun-thing-to-build, but as it gets farther along, the thought of quitting consulting to pursue it full time is pretty intriguing.

So my question was 100% about the viability of the PI as a technological foundation for this when there are other similar computers out there. If I end up finding out that some other component was a better choice, I'll still have had fun building it on the PI, but always better to have eyes wide open.

Haha, I'm glad that I came across as a marketer though. That's a first...

[davidcoton]

Try a SWOT analysis:

What are the Pi's Strengths for your project?
What are its Weaknesses?
What are the Opportunities you see?
What are the specific preceivedThreats?

If you can answer these questions,folk here may be able to confirm or rebut your answers. However, some of it may need a different audience to protect commercial confidentiality.

More generally, why do you think a commercial project based on a Pi might be crazy?
Read on this forum other people's questions about using the Pi commercially. If you are going to sell a product that includes a version of Linux, can you cope with the licencing terms, including making the source code available? (That doesn't mean that your code has to be open source.)
Have you got or can you get the financial resources to start a business (cash flow forecast)?
Do you know your market well enough?
Of course you must also have a good product, both the software you provide and the overall package. But even then, I've seen good products, appropriately marketed to an obvious and timely need, that have not flown. Can you afford that, financially and emotionally?

[piglet]

Another questions is around supply of components. While I don't expect the Pi itself to be a problem, there has been some question of the continued availability of the current camera sensor. Having read between the lines of what has been said on this forum I have the feeling that there might be a new one not long off an announcement.

[ame]

Are there any barriers to entry?

i.e. once this product is on the market and people see it is an off-the-shelf Pi with an off-the-shelf camera, what stops a competitor from creating something similar? Or recreating it as a smartphone app?

[emma1997]

Another questions is around supply of components.

This is an important point. I can't imagine basing a business on this considering the single source nature and vagaries of semiconductor industry. Some models/versions already phasing out. Hobby yes, academic maybe, but business? Kinda risky IMO.

[zcapozzi]

Having read between the lines of what has been said on this forum I have the feeling that there might be a new one not long off an announcement.

I didn't realize this. Your points on supply of components is a very good one. On the camera, are you suggesting that they are going to have a new camera soon? Or that they are going to stop manufacturing the camera?

As for competitors, this concept would provide some computer vision and ball tracking for the golf driving range industry, so not necessarily a sexy industry to be involved with. Hoping that dampens the number of competitors a bit. There are multi-million dollar competitors out there (i.e. TopGolf), so hoping that I can position this as the no-frills cheaper version of that concept (minus the bar/restaurant).

It's been an interesting experiment to think about this stuff. And I love golf, so more time at a driving range is never a bad thing.

[piglet]

See thread "OV5647 Discontinued"

[zcapozzi]

Thanks for sharing. That is helpful. I agree that having it toast bread would be fantastic. Will be interesting to see what they come out with next. I hope it doesn't have too many bells and whistles to be honest. I need a sensor that works in decent to good light and has clear colors, that's about it.

But to all your points, having the single source of components is a bit scary with the RPF is more of an educational provider than a supply chain cog. Ah well, as long as there is a board that has a camera that can run linux, it should be ok.

[liz]

There are hundreds of businesses using the Pi as their core technology: although we're an educational not-for-profit, we make and sell Raspberry Pis and add-ons into a lot of embedded business, and the device and add-ons are designed to be used in those environments.

Selling Raspberry Pis is how we make money for our educational work: it HAS to be a good, solid, dependable piece of hardware for business in order for us to be able to carry out our educational mission. You don't have anything to worry about.

[zcapozzi]

That is fantastic Liz. And your point makes a ton of sense. That does give me comfort in continuing down this path. And eventually, I'd love to be able to have some of this stuff double as a way to get some local kids involved with PIs, so the dual-purpose is fantastic.

- Zack