Research Microsite 2021

Never before have there been as many people on the road as today. That means there are more people that might get hurt in traffic. According to VRT NWS, despite corona, there were more than 200 people killed in traffic accidents this year^[1]. These 200 people will never return, and their families are broken forever. We want to find a way to reduce this number drastically.

We took a look at the data provided by imec^[2] and concluded that safety and mobility are the areas that people expect the most innovation and solutions from currently.
Naturally, our research question became the following;

How can we use new technologies to improve safety in traffic for everyone involved.

The process of our research was documented on this site. You can find all data here, or via the links below.
On the about page you can read a little more about the researchers.

After we defined our problem definition and research question, we started brainstorming with a few design principles in mind.

Design Principles #

We defined a few design principles grounded in the data and conclusions we formed from it.

• Human Centered
  During the whole process, the user should be central. This solution will be used in cities and will become an important part of people's daily lives. It has to keep the user in mind.

• Seamless
  The solution should be seamless to implement and switch over to. The users shouldn't have to adapt to the product, the product should adapt to them.

With our brainstorm done, we started evaluating our ideas, which ones were possible, which ones fit the design principles the most, and which ones fit all the criteria.

There were many ideas, from data privacy to virus detectors outside — a bit like a weather station, but for viruses —, to a projected pedestrian crossing with animations.

We concluded that updating our street lights — making them 'smart' — would be the way to go. It fit all the criteria and would offer lots of possibilities to research and actually develop.

Taking a page out of the book of 3D rendering, we can try a physical, real life implementation of a technique know as culling.
(Technical explaination^[3])

In the 3D world, culling is the term for not rendering what can't be seen. Think of a car driving behind a house, the car won't be rendered behind the house which saves computing power while the user sees no difference except a major performance increase.

Trying to apply this concept to real life — only turning on lights wherever people can see them — however, is quite a challenge. It's not possible to just turn off what's not in sight for people. We can however try and take the general concept and turn lights off whenever people are not around.
This isn't a new idea, motion sensors have existed for 71 years now, but those are pretty 'dumb' lights by today's standards. This is why we combine street lights together with 5G to make them smart, save even more energy than with motion sensors and enable safety features that weren't possible before.

From a driver's point of view, you can only see a few cars ahead while you can see street lights for hundreds of meters in the distance.
We would also use the street lights as an early warning system to alert drivers that an accident has occurred and that they should slow down.

User Testing #

Of course we also conducted user testing. To gather some insights of how our product is received by the users, we made a prototype and gathered people around to test it out.

Data can be found here.

Conclusions #

After testing our prototype we collected some important insights.

The short light range prototype resulted in a far less enjoyable and safe driving experience in comparison to the long light range. This means we should definitely continue the concept with a longer light range.

We also noticed that the lights snapping on was annoying, and distracted from the actual driving. We should definitely give the lights a fade, so people on the road can stay focused on the road and aren't distracted.

During post-test interviews we talked about 5G and we only received positive answers about this. People don't seem to be bothered about data processing. It is definitely the way to go for this project.

5G Specification #

As concluded from our user testing, our users did not have any issues with the data being sent to the street lights, as long as they were not being tracked, the data wasn't being stored, or used to issue speeding tickets.

Looking into the 5G spec, we would use the URLLC spec^[4] which "is a new service category in 5G to accommodate emerging services and applications having stringent latency and reliability requirements." (Ji et al., (2018).) Or, in short, exactly what we need.

URLLC defines that the user plane latency target is set at 0.5ms^[5] (paragraph 2, line 4). This is insanely fast and would allow for direct p2p communication to the street lights without needing a server inbetween processing, or storing the data.

This greatly improves user privacy while also guaranteeing reliability and uptime.

5G also offers a major increase in bandwidth, up to 20Gpbs^[6] peak data transfers.

These are the reasons we opted to use 5G instead of current standards like 4G or 3G. Compared to those, 5G is much faster, more powerful, and has better standards — although they are still standards, not actual results yet — than 4G and 3G.

This is where our second design principle comes in.

In order to have cities adopt this technology, we will need to get the upfront cost of installing these new lights as low as possible.
In a best case scenario, we'd also make it as easy as changing out a light bulb.
This way it becomes super easy for big, but also small cities to adopt this technology, which results in a faster and larger rollout.
Cities can also choose to do small(er) trial runs before making big commitments to their budget when the upfront cost isn't enormous.

Conclusion #

In conclusion we have found that our concept can definitely work. User-testing has shown that the lights help with maintaining a safe speed, while giving the drivers enough light to see upcomming obstacles. It should greatly enhance safety.

5G is an amazing new advancement in technology that will be at the core of our project. According to our interviews people are not worried about data processing — given it's solely used as required in order to operate the lights — and are happy with the fact that tech of tomorrow is being used in their smart-cities today.

Further testing should be done on a larger scale. Reactive simulators can give us even more reliable results, and insights for further improvement. Funding would be needed to conduct these more expensive, but far more realistic tests.

Sources #