LabVIEW - Looking for some Figures

[wiebe@CARYA]

---

@labview-crash wrote:

So you've developed your product and now you're wandering if there is a business case?

I thought it worked the other way around.

---

It might not work either way 😕.

Seriously, it's hard to make money from a LabVIEW product.

---

@labview-crash wrote:

Now as a joke, if you believe in AI, ask your questions to ChatGPT 😉

---

😁

It (he? she? they? is there a term for an AI yet?) won't have recent info though.

[wiebe@CARYA]

---

@Taggart wrote:

Ask Emerson.

---

Yes, they didn't get those numbers either...

[AloisLIEN]

Actually, at the beginning, we started to develop HAIBAL for an internal need. There was no suitable tool in the LabVIEW environment that met our needs, and logically, we assumed that we were not the only ones in this case.

When we started talking a bit about HAIBAL, some scientists expressed their interest in the product. We put the library on sale in early 2023, and we have already sold licenses (not 100 of course but we are just starting). We are over-motivated, we release a new version of our beauty almost every week, and for now, we are sticking to our ambitious roadmap.

I sincerely believe that there is room for such a tool. AI is increasingly present in all sectors of activity, and yet, there is not yet a hyper-complete tool that allows LabVIEW users to "practice" AI simply and efficiently. According to Accenture, 80 to 85% of projects integrating AI remain stuck at the PoC stage. One of the reasons? The complexity of AI development environments. HAIBAL solves this problem.

Of course, this is just my opinion (that official figures from NI would help me to consolidate :p).

PS: I agree that it may well not work, but as I always tell my young Princesses: "Only those who create nothing never fail." So let's try!

[santo_13]

---

@AloisLIEN wrote:

Actually, at the beginning, we started to develop HAIBAL for an internal need. There was no suitable tool in the LabVIEW environment that met our needs, and logically, we assumed that we were not the only ones in this case.

When we started talking a bit about HAIBAL, some scientists expressed their interest in the product. We put the library on sale in early 2023, and we have already sold licenses (not 100 of course but we are just starting). We are over-motivated, we release a new version of our beauty almost every week, and for now, we are sticking to our ambitious roadmap.

I sincerely believe that there is room for such a tool. AI is increasingly present in all sectors of activity, and yet, there is not yet a hyper-complete tool that allows LabVIEW users to "practice" AI simply and efficiently. According to Accenture, 80 to 85% of projects integrating AI remain stuck at the PoC stage. One of the reasons? The complexity of AI development environments. HAIBAL solves this problem.

Of course, this is just my opinion (that official figures from NI would help me to consolidate :p).

PS: I agree that it may well not work, but as I always tell my young Princesses: "Only those who create nothing never fail." So let's try!

---

What fraction of the AI users use LabVIEW as their choice of language?

How does HAIBAL compete against mainstream tools negating the easy-to-get-started part?

[TiTou]

One thing you can get from NI, even from their website, is the list of system integrators who do vision.

These companies are potential clients for your product, as you are based in France, I suppose you have already identified them, right?

https://www.ni.com/fr-fr/search.html?pg=5&ps=10&sb=%2Brelevancy&sn=catnav:ptr&q=vision

[AloisLIEN]

---

@santo_13  a écrit :
What fraction of the AI users use LabVIEW as their choice of language?

How does HAIBAL compete against mainstream tools negating the easy-to-get-started part?

---

What fraction? For now, I can only suppose this fraction is tiny, since there is no efficient AI tool on LabVIEW.

How to compete? The banal answer would be "we make the best product", but that wouldn't be enough to fight against giants like Google and Meta. Concretely, we will try to unblock all the companies/labs who are stuck in their AI projects because of the complexity and/or their difficulty of recruiting suitable staff (on this type of customers, we already have some very interesting leads, and I hope to be able to announce great partnerships in the coming months). We will then prove to them, factually, that HAIBAL is out of the ordinary. We are clearly not going in a sprint but on a marathon, hence my search for investors.

[AloisLIEN]

---

@TiTou  a écrit :

One thing you can get from NI, even from their website, is the list of system integrators who do vision.

These companies are potential clients for your product, as you are based in France, I suppose you have already identified them, right?

https://www.ni.com/fr-fr/search.html?pg=5&ps=10&sb=%2Brelevancy&sn=catnav:ptr&q=vision

---

You're right, and yes, I already identified these companies. So far, I just haven't had time to contact them. I will do it asap.
Thank you for the idea anyway!

[wiebe@CARYA]

---

@AloisLIEN wrote:

I sincerely believe that there is room for such a tool. AI is increasingly present in all sectors of activity, and yet, there is not yet a hyper-complete tool that allows LabVIEW users to "practice" AI simply and efficiently. According to Accenture, 80 to 85% of projects integrating AI remain stuck at the PoC stage. One of the reasons? The complexity of AI development environments. HAIBAL solves this problem.

---

You're competing with (for instance) OpenCV.  OpenCV has at least some AI tooling. OpenCV can, and is, called from Python from LabVIEW...

It's hard to compete with free and established tools, even if your tool is better...

As a spin-off of an internal project, I'd go for a quick time to release, because there's 0 proof of the ROI you'll get, if any. If the product isn't release ready, be honest about it and sell it (for less, or even give it away (as trial)) to early adapters. Go from there...

Keep maintenance, service and sale costs in mind. These are expensive.

We found that selling our tools is more expensive than giving them away (and get some exposure from it).  Or at least, we haven't found a product that contradicts this yet.

[TiTou]

I very much agree, you'll probably get more business out of it if you release it for free/trial - even open-source -, and be confident that you can sell engineering to those who don't have enough expertise in-house to make it work in a real-world application.

[Taggart]

You might talk to Jim Kring. He might be willing to give you some numbers around VIPM package downloads. It's a proxy but might be useful.