How Data Will Drive The Next Big Design Trends

One trend I’ve been noticing in the design world is a heavy emphasis on data-driven design. I think it’s going to be one of the most important industrial design trends of the 2020s and beyond. This video will give you a better understanding of why data will be so important to designing the products we make, the ethical controversies around data collection, and why some types of data are much more valuable than others.

0:00 Intro
0:26 What is Data?
0:56 HoloLens 2 & Ergonomics Data
2:04 Digital Twin Data
3:42 Data and Diminishing Returns
4:26 Data as Competitive Advantage
6:25 Czinger, Nike, & Performance
7:09 Shein & What To Make Next
8:30 Data & Privacy
10:12 Key Takeaways

All content written and edited by John Mauriello. John Mauriello has been working professionally as an industrial designer since 2010. He is an Adjunct Professor of industrial design at California College of the Arts.

CREDITS:

[nonameishere]

You're born to do this 🙌 one of the best speakers. I have a short focusing span but your videos keep me engaged from start to finish every time

[capnthepeafarmer]

Great video! I appreciate your insight into this data design! I work in an industry where "data" is thrown around a lot and I hope I can help make some distinctions and examples I've found in my industry. One distinction is the difference between "Big Data" "normal data" and "simulations". That's Big data with a capital "B", and the distinction there, as you touched on, was how big it is. Like, for example when most people think of data they think of files of text or spreadsheets, and for reference the maximum file size of an Excel spreadsheet is about 2GB, your typical file is less than 10MB. That's a lot of data. But "BIG" data is many orders of magnitude higher and is often updated in real time. I think there is a science mapping system that intakes about 200GB of data a DAY. That's "BIG" data. It's incomprehensibly large. And in essence, big data is queried to look at trends and help make inferences and predictions.

Normal data (my personal definition) is any data set which can be analyzed on a single computer system. So in the <2GB range which any data scientist could pivot table, or use minitab or M to analyse. In my industry what's thrown around a lot is "Industry 4.0" ( I4.0 ) and "Big data" and it's more or less just buzz words the leadership likes to make it sounds like work is being accomplished. What it really comes down to is a very specific set of collected data that a single person on a single computer system analyzes. Think of a vibration sensor on a motor, you just monitor the vibration amplitude to help predict a maintenance interval. It's a very mundane use, although the data would fall under my definition of "normal" data.

Simulations do generate a lot of data, but it's distinctly different than "big" data in that it's the generation of the data, and iteration to a goal. The lattice simulation you mentioned requires certain boundary conditions that are inputs and it's given a goal to seek, i.e., your output. The user gets an output of "this is the optimum design for your input conditions". Which after a lot of the hype has died down is that it's just another design tool and they are error prone, it's not replacing anyone. Like, a generative bike frame design that uses the least amount of material may not actually be a good design. I'm generalizing from my experience, generative/parametric, "neural" AI, etc designs are usually mediocre because it's often times trying to optimize the wrong goal, or the data available is incomplete or there are omissions that the designers didn't consider. Lets take a look at the generative bike design that tries to minimize material and use aluminum or 3D printed metal because less material is "sustainable" and lightweight. Say the output is a very intricate 3D aluminum structure like you showed. The experienced designer would have just made the bike from carbon fiber and saved 50% on weight and material without doing any work. So in my experience generative design gets a bad rap because it's used too early in a design process and the designers try to optimize the wrong thing.

[SJ-wdrz]

The production value of the videos really shot up, now you have really strong intellectual themes with great visuals. Really enjoyed learning about we can and should use data for the benefit of creating better products.

[WillGibbons]

Nicely done, John! I think your videos are getting better each time!

[christopher]

This is totally the future of industrial design. I would love to learn how to use simulation software someday. I'm really fascinated with aviation and I imagine they've been doing that for a while in their industry.

[MrDecaliostro]

I use big data analytics in my branding designs, works like a charm!

[awogbob]

Video could easily have 100x the views. Great quality keep it up

[joe_fabricator]

Found your channel thru Derek Elliots channel. Great stuff. Thanks John.

[Bird]

My daughter is mildly obsessed with Shien. After hearing this I understand why. Great work on their part!

[McRyach]

Thank you.
Great video.
I love how you're making a disclaimer about ethics of a company that using a data. It is definitely a smart way to design. But that design objective could be unethical in itself.

[soumyapande]

Love these type of videos! Can you do more about using generative design in product design

[EmilioBaldi]

4:31 for some things you don't need IA: years ago an Italian nurse invented a machine that counts up the equipment dispensed and counts down the disposed. 100% accurate.

[fimolo]

Great video! Love the way how you also reflected the critical part of Shein and data collecting.

[pranitpopli]

Noice!, I remember shein getting banned in India cause of data privacy.

[OmDahake]

can you make a video on design University like what to know, what to expect, how to deal with pressure or if it is even needed or some alternative ways and how you did it and how you survived and got your foot in the industry and became what you are now ?

[hizzaddinno]

2:49 Instead of working against the heat generated by the visor, try linking the room's AC system to the visor's system so that you can emulate fake sunshine heat and wind on the user's face.

[hypersonicmonkeybrains]

For the HoloLens there should be an option to scan your own head and 3D print a custom headset case to fit your own head.

[TheBigjoao]

Really like your videos! Thanks for making them. Here’s a question for you, do you first make the script and them edit the video or compose the video with things you want to say and then voiceover the whole timeline?

[ihx]

honestly data in general should be accessible to anyone caus its a product of everyone

[sshlokmishra]

Very late for this but awesome video 😁...