Shaping the future of data-enabled design research with Data Foundry

Math­ias Funk from Eind­hoven Uni­ver­si­ty of Tech­nol­o­gy joins forces with 4TU.ResearchData to improve the FAIRness of design data.

Math­ias Funk is an Asso­ciate Pro­fes­sor from the Depart­ment of Indus­tri­al Design at TU/e. As part of the Future Every­day group, his research explores the inter­play between emerg­ing tech­nolo­gies and design­ing for people’s every­day life. 

Math­ias is focussed on using data in design, and par­tic­u­lar­ly, ‘data-enabled design’; an approach that involves the remote cap­ture of data about user behav­iours and pref­er­ences using sen­sors. Such con­tex­tu­al data can be used to inform and inspire the design process, and to cre­ate and val­i­date design inter­ven­tions.

Data in mod­ern design 

Math­ias empha­sis­es the impor­tance of data in mod­ern-day design and design research. “Our world is becom­ing increas­ing­ly more tech-dri­ven. As humans become plugged-in to the ‘Inter­net of Things’ for their dai­ly rou­tine, there’s a grow­ing demand for devices that can process and gen­er­ate data in a high­ly sophis­ti­cat­ed man­ner.”

“Such intel­li­gent devices can have embed­ded sen­sors and capa­bil­i­ties to seam­less­ly exchange data with oth­er devices and ser­vices. They may rely on com­plex algo­rithms that encode the rela­tion­ship between inputs and out­puts,” says Math­ias. 

He con­tin­ues to explain that these devices remain a ‘black box’ for many. “Design stu­dents often con­sid­er arti­fi­cial intel­li­gence and machine learn­ing to be like a form of mag­ic sauce. They under­stand that inter­ac­tive devices use and pro­duce data, but what hap­pens in between remains a mys­tery to them.” 

There is a lack of under­stand­ing of how devices process data and what is pos­si­ble when data is used in design. This is large­ly due to tech­no­log­i­cal and method­olog­i­cal bar­ri­ers in col­lect­ing, pro­cess­ing and rep­re­sent­ing data in a design, and the lim­it­ed pres­ence of FAIR data in design dis­ci­plines. Tra­di­tion­al­ly, design data is qual­i­ta­tive, dif­fi­cult to scale-up and stored in pro­pri­etary ways mean­ing that data is inac­ces­si­ble to the research com­mu­ni­ty.

Shap­ing the future of design research

Deter­mined to make a change, Math­ias is devel­op­ing Data Foundry with a small team at the Indus­tri­al Design depart­ment.

Data Foundry is an online infra­struc­ture for pro­to­typ­ing and design­ing with data. Researchers can col­lect, store, process and export data from var­i­ous remote devices in real-time.

Math­ias hopes that Data Foundry will improve data lit­er­a­cy, the FAIR­ness of design data and help researchers to under­stand how intel­li­gent devices inter­act with and through data.

“Data Foundry offers design researchers a unique oppor­tu­ni­ty to learn about remote data col­lec­tion and pro­cess­ing beyond the low-scale qual­i­ta­tive,” explains Math­ias. “Researchers can inter­act with their data as soon as it is cap­tured and, there­fore, gain a bet­ter under­stand­ing of a con­text or a designed expe­ri­ence.” 

Math­ias adds that Data Foundry also serves as a research data man­age­ment sys­tem that helps to pro­mote data reuse. “Data from a vari­ety of sources is col­lect­ed in a com­mon, uni­fied for­mat using over ten dif­fer­ent dataset types, and is described using meta­da­ta, mak­ing it eas­i­er for researchers to under­stand, com­bine and reuse each other’s data.” 

He hopes that the uni­fied infra­struc­ture for data col­lec­tion will encour­age a trans­par­ent and col­lab­o­ra­tive way of work­ing where­by stu­dents, edu­ca­tors and researchers build trust and devel­op cre­ative ideas togeth­er. 

A col­lab­o­ra­tion with 4TU.ResearchData 

Since its incep­tion in late 2018, Data Foundry has been pilot­ed in var­i­ous Indus­tri­al Design cours­es, includ­ing the Data-enabled Design Master’s degree course, and is to be part of the cur­ricu­lum of upcom­ing cours­es on data, sen­sors, and machine learn­ing.

Cur­rent­ly, Data Foundry sup­ports more than 250 design projects at TU/e. Whilst infra­struc­ture com­po­nents still need to take the test of time, the next step is to ensure that finalised datasets from closed projects are trans­ferred to a trust­ed data repos­i­to­ry for their long-term preser­va­tion, access and shar­ing. Math­ias has decid­ed to join forces with 4TU.ResearchData to make this hap­pen.

The objec­tive is that once a project is closed in Data Foundry, its data and meta-data can be auto­mat­i­cal­ly import­ed to 4TU.ResearchData, an inter­na­tion­al repos­i­to­ry that boasts more than 8,400 datasets in sci­ence, engi­neer­ing and design dis­ci­plines. Datasets will be assigned a DOI and described using rich meta­da­ta to make them more acces­si­ble to the wider research com­mu­ni­ty. 

This year, tech­ni­cal team mem­bers from Data Foundry and 4TU.ResearchData will work towards estab­lish­ing stan­dards for inter­op­er­abil­i­ty between the two infra­struc­tures so that the exchange of finalised data and meta­da­ta is easy, safe and effi­cient. 

Math­ias believes that many heads are bet­ter than one when it comes to shap­ing the future of design research. 

There are thou­sands of flavours of FAIR data and it’s impos­si­ble to tack­le them all work­ing in iso­la­tion. With 4TU.ResearchData, a con­sor­tium ini­tia­tive of three Dutch tech­ni­cal uni­ver­si­ties, we can cap­i­talise on the strength and knowl­edge of many experts. Let’s see how we can share data and do more.

As the new year brings excit­ing prospects, we look for­ward to keep­ing you updat­ed on the lat­est devel­op­ments from Data Foundry and 4TU.ResearchData. If you’d like to learn more about this col­lab­o­ra­tion, get in touch. To learn more about Data Foundry, head to the doc­u­men­ta­tion, use-cas­es and Devel­op­ment Blog. Or, see Math­i­as’s pro­file.

Writ­ten by Con­nie Clare (4TU.ResearchData)
Cov­er image illus­trat­ed by Con­nie Clare (4TU.ResearchData)