theuttick

I am curious as to why you may have been rejected. Seems like a decent idea and you guys are programmers. Do you have a MVP yet?

Wow. I am very embarrassed about that. I appear to have broken something recently. The bug report and feature request are working again.

I am also in the process of moving from a lamp stack to python/Django on AWS. I'll have to look at the ssl and see what's happening.

There would be both logistical and technical benefits.

1. A direct link between the CAD, the math, and the documentation. I don't believe that Solidworks will update its models without pulling the matlab output onto the server and the documentation won't update without creating the PDFs and pulling them into the PDFs.

2. CADWOLF will allow you the ability to create template items like fittings or trusses that consist of a CAD model and the math. Your junior engineers drop this into their designs and give defined inputs like loads and the truss redesigns itself.

3. The part tree system organizes all of these facets and gives you the ability to see real time system weight/cost/whatever instead of having your people update spreadsheets.

4. A link between your work and neighboring parts/systems. Although the stack you described "works" whenever the jackhole sitting in the cube next to you updates his model, you have to wait for him to tell you and then go through the entire redesign/upload/redo process. CADWOLF will see those changes and update accordingly.

5. One system vs Matlab, Solidworks, C#, PDFs, LaTex, etc. Since it is browser based, it can be seen by anyone/anywhere. This is more logistics, but it prevents you from having like six pieces of software running and it prevents you from making a mistake somewhere.

6. Scale. We will have an AWS server side that can be scaled up instead of being limited to your desktop. This will also help small companies that don't want to buy large computing systems.

7. Process. Since you have access to every person's documentation for a system through a part tree, you can open that doc and check their work from anywhere. A checkoff system will eventually be implemented.

I did something similar to what you are doing with all of the different pieces of software. It really doesn't seem that bad until you realize that it could all be done automatically. There are also far fewer engineers capable of doing what you are doing than most people realize. You're like in the top 10% of production in my experience.

Note about discussion on submissions - the comments closed at the standard time so when people looked at old submissions they couldn't participate in the discussion.

EE problems will be addressed in later versions of the system. The hope is that it will be able to monitor voltage and current requirements and tie those into the math. Circuit board type problems are a back burner thing for now.

If you want to contact me, you can get my email from the team page on cadwolf.

Hi. Really appreciate the mention.

This is indeed a huge bite, but it has the legitimate potential to change the way we design and build everything. The comments on the original thread are closed, but I'd be happy to answer any questions in this thread.

The platform has 4 components. One component mimics a folder tree and shows all of you documents and subfolders and stored images.

The second is the documents themselves. Here, you can create a document and add text, equations, plots, etc.

The third is what we call "datasets." A dataset lets you import an excel file, a web page, a text file, or just paste data and store the parsed results. You can then import that dataset as a variable into a document.

The fourth is the part tree module.

Everything other than the datasets is done from within the framework. You can paste text into a document if you like and work with it like a word document.

Can you aim the cannon? If not, you would be limited in the orbits that you could access without some sort of on orbit propulsion. Is that true?

Thanks to everyone that has asked a question. I wanted to add a little more info on how this will speed up engineering in the future.

A "fitting" is a small component that usually attaches a rod to another item. The fitting consists of a base plate and two vertical extrusions. There is a bolt at each corner of the base plate connecting it on that end and an eye bolt sits between the two vertical extrusions and screws into the rod.

I CADWOLF, you will be able write a single web page or "document" that takes in the load on the fitting through the rod and then calculate the base plate size, the thickness, the size of the vertical extrusions, and the sizes of the bolts needed. These items will populate your part tree automatically. When the load changes, everything updates automatically. If you change this document, the design updates.

Now imagine that you are building a large structure with dozens of these fittings and any engineer can simply place the fitting into their structure in the same way that you call a function. You don't have to worry about whether or not the engineer did the design properly as everyone is using the company standard analysis and the engineer can go about doing the more complicated things.

Now imagine that CADWOLF provides a large array of built in structures like this where a single engineer can design large and complicated structures faster than teams of engineers do today.

Pretty much. The CAD will be fully browser based and connected to the documentation and analysis side as well.

There are a lot of things that are somewhat similar at first glance. However, many of those require the user to write a document of some kind in a text editor and then upload it to a server where it is viewed through a URL as a web page. With CADWOLF, the user creates a new page and edits it in a WYSIWYG format.

You are very right in that a product in this space has to meet a very high bar before it is useful compared to other things like Matlab.

I have functionality to do things like differential equations, linear algebra, FFTs, but not really hard math things like symbolic integration. So, right now my targets are users that need something more powerful than excel but don't need the higher level features of matlab or python. To help acquire those people, I am writing a server side on AWS to let them solve large problems and simply pass the cost they incur on to them (for now). In this way, I can target small companies and let them solve some large problems without having to buy large servers.

I am hoping that this side can pay for the development of the CAD system.

What sort of simulation would you do that has not been done at Boeing, UW, etc.?

I am an aerospace/mechanical engineer. I think that the manufacturing aspects of this would be far more difficult to overcome than the mathematics of proving the idea as viable. In other words, it works in theory but in reality you can't make a tube with smooth enough connection points etc. Will you look at that too?

If you get this and are looking for minions let me know.

There is no integration with existing tools. There are some very good CAD systems out there - Catia, ProE, etc. However, for security reasons, there is no way I could link up to that software on the server.

Hi,

I am an aerospace engineer that worked on the Orion capsule in Houston as part of the constellation program. I have a BS and MS in aerospace engineering from the University of Texas at Austin.

We use MS Word mainly because it's there, everyone has a license for it, and everyone just accepts that we use one tool for documentation and others for analysis. We can change that.

The time savings comes this way:

Right now, you design your components in Excel, Matlab, etc and then document them. When a requirement or neighboring system changes, each engineer must then go back and make the necessary changes to their math, their documentation, and their CAD. This can take weeks or even months for small changes as each engineer has to go back through their work.

With our system, the user designs his component as part of the whole system. When a requirement or neighbor changes, the software filters those changes out and the components redesign themselves. This will be true for the CAD as well. Weeks or months of work for dozens of engineers or more then takes seconds.

A very, very good answer. However, in this case the diversity the man at twitter should be looking for has everything to do with a person by person case and nothing to do with race.

Yes, white and black people will likely have different experiences growing up. Most white people that I know - including myself - were routinely attacked at school because of their race. Every white person that I know has had to look at a job interview and wonder if he didn't get called back because the job was only for certain people.

However, none of this would help the white person build a content moderation feature or prevent abuse or protect freedom of expression better than the black person.

May I ask specifically how a person's belief that their name was the cause of failing to get a job leads them to create a better content moderation feature?

The one thing that he never explains is exactly what "diversity" he is looking for at Twitter. It seems that the only thing that matters to him is the color of the person's skin. In other words, an African-American MIT CS grad that grew up in Iowa is somehow going to come at problems differently than a white guy from the same town and school and with the same background. I wouldn't really call that diversity, that's just racism in that one man's view on a technical matter is somehow different because of his race.

What they should look to do is recruit people who got to programming or designing through paths other than MIT or Stanford. Hire the guy that was a painter and then started building website UIs when he saw how those skills transferred. Hire the guy that graduated from the small state school and spent all of his free time web programming. That - at least to me - is the kind of diversity that comes at problems from various angles.

The VP idea with names was stupid, but if I walked away from a job every time one guy had a stupid idea i'd be locked in a closet somewhere howling at the world's stupidity.

My 2 cents. I got to get back in my closet.

I recently went went public with something similar - www.cadwolf.com

Mine is similar in that it does math in the browser. However, the goal of cadwolf is to do math and documentation on the browser and remove the need for the user to know a programming language.

I really like what you guys did, but why would someone use blabr to run coffeescript instead of just running their own program? What is the use case outside of a learning tool?

Also, I would love to know what you think of cadwolf.

Sounds good. If you're doing space stuff in houston, I likely have drank with someone you're working around. Be interested to know what you're working on.

I agree. There is a huge amount of "inertia" in the engineering field. People find something they like and then use it forever.

Thanks for the input and I totally get what you are saying. It may help to know that the end goal is to tie in a CAD system. This way, each "document" contains the math to design a component while the CAD reads from that to display it. This way, the CAD model automatically updates with changing design requirements just like the part trees already do. What you see is really just a first step.

Interesting that you say that. There is a spot on YC applications that asked if you had any other ideals that you may have applied with or something along those lines. I mentioned pivoting to just that sort of thing.

I like those projects. What I am doing differs in purpose and function

Those platforms require some sort of installation and are usually meant to be tools for a programmer to create a web page that utilizes existing code like python, etc. My platform has no coding in that sense. You create a new file by clicking an icon and add blocks of text, equations, etc through a graphical interface. All the user would enter from a code standpoint would be something line "a=sin(10x)+cos(3x)".

Whereas Jupyter and others are tools for programmers to use, my platform is a tool for other people to solve and document problems. They target programmers whereas I target engineers and scientists. With them you are writing code and with me you are using a word document that also solves equations and makes charts.

There actually is an example document that any user can fully edit. It's under the docs and examples link.

I have applied several times, including this round. I am a single founder and I understand hesitance to take people in that situation. I am trying to recruit some co-founders but it has been difficult finding the right people.

Matlab, Maple, etc are very good at what they do, but they have two problems that my platform solves.

1. Your manager and/or customer can't see your work unless they have a license and can understand the language. Mine only needs a web browser.

2. If the analysis changes, the documentation must be updated as well. Mine does everything on the same web page.

As for the questions, those are all ones that I have asked myself about several times and here is where I am now.

(1) It is not open source and you really are kinda locked into using the system once you get going. I really couldn't find a way around this as the real appeal for most engineers is the absence of the need to code stuff. I have documentation that describes how I solve most algorithms and some demos.

(2) Scraping web pages is high on my to do list. You can paste most forms of data already, but you can't yet import files directly. Seeing that this is high on someone else's list I may put most of my effort there.

(3) Right now, everything is done locally in javascript but there is no way to save changes without server access. Part of my business plan is to use a server side code to solve large data sets and possibly charge for that. I have been surprised at how well the browser has handled most problems though.

(4) The main problem that this will solve for most people is this ... You get data from a test run somewhere. You put that data on a server, run a Matlab or Python code on it and then create a word document to pass on that info. With my system, you upload the info, analyze it, and display the results all on the same web page. It isn't just about analyzing the data, it's about doing everything on the same platform.

(5) The plotting functions are pretty easy to use and you can get something up in a few seconds. I also have 3D plotting.

Submitter here,

I am a long time everyday reader of HN but I have never posted or really commented. I am a structural engineer with a few years of work in the space program. I have spent a great deal of time building something that I think can redefine the way we do engineering and I am close to finished with an initial version.

I am extremely nervous about posting this, but I would love to hear what everyone thinks. (There is a video in the overview section)

Thanks