A new tool predicts when users will reject a new technology

There is something paradoxical about our relationship with technology. We have very high expectations that new technology will solve the greatest challenges of our time. At the same time, we are quite often skeptical about using new technological solutions.

When new technology fails, it’s not always because it does not work as intended. Sometimes, people simply don’t want to use it. One researcher believes this should be predictable.

Sarang Shaikh, a Ph.D. research fellow at NTNU in Gjøvik, believes that being able to predict which of the two scenarios will play out is important.

“Not being able to predict whether people will adopt new technologies can result in significant losses of both time and money,” he says.

Shaikh and his colleagues have now developed a tool designed to prevent this from happening. The work is published in the journal Data in Brief.

Expensive border control technology has not been a hit

Shaikh and his colleagues were tasked with investigating why new technology recently installed at airports and border crossings across Europe is so underused.

Travelers are accustomed to having a person check our passport when we are on our journey. With the new technology, people get channeled through a set of barriers. Your passport and fingerprints are scanned, and a computer compares the photo in your passport with your face. If everything is in order, the barrier opens and you are allowed to pass through to the other side.

The EU invested millions of euros to automate border control. Several years after the technology was installed and made available across much of Europe, many travelers still choose not to use this expensive technology.

“It is hard to imagine anything simpler and more efficient. Why, then, do so many people still prefer manual checks?”

The EU Commission therefore asked researchers for help to understand why this was the case—and how to avoid it in the future.

It’s about more than just technology

“Before we could develop a tool to predict whether a technology will be used or not, we had to understand which factors are most important in determining whether someone chooses to use a specific technology or not,” explained Shaikh.

And it turned out that it was about much more than just the technology itself.

By interviewing users of automated border controls, as well as the border guards who operate them, the researchers identified three crucial factors that determined whether travelers would switch to using the new technology:

• User profile: How old is the user? What sex are they? What level of education do they have? Is the person a seasoned traveler or a first-time traveler?
• User perception: How did the user feel after using the technology? Was it a positive experience? Or did it cause negative feelings like stress? Did the person use the technology in a good way or not?
• The surroundings: What is it like when the traveler decides to use the technology? Is there a long queue? Long waiting times? Does using the technology seem simple and seamless, or does it feel like a cumbersome process?

Huge volumes of data from social media

In addition to interviews from five pilot studies around Europe, the researchers gained a lot of valuable data from online discussions.

“We also collected and analyzed large volumes of data from social media. We then compared the data with what we had already gathered from the interviews. We found that what was discussed online largely aligned with the experiences reported by the border guards,” said Shaikh.

“The big question was whether we could generalize the results to also apply to technologies that do not yet exist or have not been rolled out,” he said.

Making data available to other researchers

The goal is for anyone who wants to develop or introduce a new technology to have access to a tool that allows them to assess how the technology is likely to be received.

“Based on this information, they will be able to make more informed decisions about whether or not to move forward with the technology,” said Shaikh.

To test whether this was even possible, the researchers used machine learning to train a model based on the data they had collected.

The model has already been tested at several automated border crossings across Europe. It enabled the researchers to predict whether travelers who had never used the new border controls before would choose the automated or the manual option.

So far, the findings suggest that the results can be generalized beyond the original dataset, and the model can already predict with high accuracy the extent to which a technology will be used compared to previous methods.

Shaikh now hopes that other researchers will make use of the datasets he and his colleagues have made publicly available.

“When I started studying this, the lack of data was one of the biggest challenges we faced. We have now published two datasets that have never been available before. Researchers and other interested parties will be able to use these to train even more advanced and accurate models in the future,” said Shaikh.

The interdisciplinary research group MR PET has developed the tool. Professor Sule Yildirim Yayilgan, Associate Professor Erjon Zoto, and researcher Mohamed Abomhara were also on Shaikh’s team. Shaikh’s research is based on the METICOS project.