User Experience

[tdobson]

The workflow is backwards on mobile - you need to choose which model you're using first, and then input the Beam Length

[czmj]

It's showing up on mine correctly - will have to test more with you :)

[tdobson]

It tells me I need to add a model before a beam length, but shows the beam length first.

[czmj]

I think it's important that the Beam Length is first, because that's sort of the "setup" for the entire calculation. No matter how many loads you add, the beam length will always stay the same, and it's greyed out after you enter the first load.

Does this seem like a better solution:

1. Setup the beam properties
   • Beam length
   • If the app was extended to calculate other things that required other beam properties (eg. material stiffness), those would be here.
2. Setup an individual load
   • Choose a picture
   • Input W and alpha
   • "Add a load" button
3. Results
   • "x loads added"
   • Lists the loads already added
   • Answer - fixed end moments are...
   • Calculate button no longer needed - "add a load" could update this tab

[tdobson]

Can we write a user story for this?

http://en.wikipedia.org/wiki/User_story http://rubygem.me/2014/07/21/bringing-user-stories-to-life/

From a non-civil engineering background, I suspect one of the problems is that I don't understand who's coming to this, and from what point of view. I suspect that I may be making poor suggestions because I don't understand this well enough. :)

Perhaps the wiki would be a good place to do these?

[tdobson]

eg: https://github.com/tdobson/FEM/wiki/User-Stories

[tdobson]

I've read the user story. I'm not really sure it's clear to the user how to make use of the system.

I suspect if you were to put it in front of someone, they'd be confused about how they could it was useful to them, and what to do.

Do you know any people who you can get to use this whilst you watch them using it? Like, literally sit next to them (saying nothing) and watching them trying to solve a printed-on-paper problem you've given them? I think you'll get a lot out of watching real users using it to solve real problems - and that will help you understand their user experience challenges. :)

[tdobson]

You might give someone a problem: 1) you're designing a bridge across a 50m deep, rocky ravine for a quarry in Arizona. The gap is 10 metres, and your surveys have revealed sticking a 1m support 5 metres across is best. It's a road bridge that must be able to sustain two full $specificweight industrial dumper trucks on each span. That's quite heavy! Your suppliers can give you concrete in width a or width b. Which do you need for this bridge?

2) Your manager tells you the quarry saw the cost, and realised that the bridge was really only necessary for their staff to visit the nearest greasy spoon cafe, 100m away. With that in mind, they thought the road bridge as unnecessary, and a pedestrian bridge capable of sustaining 10 people would be sufficient. Since the load is much less, the central support probably isn't needed either)

Our suppliers can also offer us even thinner prefab concrete beams. What do we need?

My feeling is that (with questions that actually make sense) end users should understand that they have a solveable problem, and the FEM calculator can help them solve it.

Watching people trying to solve it, and watching what they do (or don't do) will be useful to see how it can be made more intuitive.

(Note: the users will think they know how to improve it - but they can't imagine things vastly different to the status quo - beware of following their suggestions in full without applying critical thinking first)

[czmj]

I expect the best way to improve the experience (although I have no idea how I'd code it) is to make the whole thing a lot more graphical. Rather than choosing a load case and entering how far along the beam it is, you could just drag and drop a load onto the beam. Something like this but with quantifiable outputs and the possibility to add more than one load.