Driverless Cars: Bridging Academia & Industry

Driverless Cars: Bridging Academia & Industry

A semester project ending in Silicon Valley? Four TANT master students share their experience and lessons learned!

Last modified: 24.08.2017

Building the future (the intro)

Regardless of how much you know about self-driving-, driverless-, or autonomous vehicles (AVs), regardless of when and in what form they will come, it is now that humanity writes the chorus of this new technological symphony and decides on standards for the future. Hence, it is now to reflect on the potential impacts, engage those being influenced and set the agenda on how we want our future mobility system to function. This is why the current academic and industrial momentum is so enormous and engages the creme-de-la-creme of researchers on AI, sensor technology, human-machine-interaction and many other areas to overcome some of the challenges that AVs and their deployment bring up. One area that needs to contribute and find a way to engage in these discussions is social sciences, to reflect on how co-living with a robotic system like an AV should and could look like prospectively.

Despite many rather general, but important, concerns like the impact on the job market, on driving experience or on the city’s mobility system, there is currently global effort put on investigating the interactional needs between AVs, normal cars, pedestrians, cyclists or motorcyclists - and how to design the technological solution of an AV communicating with other road users. In the following, four Techno-Anthropology master students share their experience on building up expertise in this research area, on their major learnings and their journey to present their findings at an expert workshop of the Automated Vehicle Symposium 2017, in San Francisco.

 

1. Become Cutting edge & find the status-quo of research

After our first semester project which was useful for building an understanding of the current situation in the development of autonomous vehicles, we had a clear vision for our second semester, namely to generate insights for the developers, and make a first step into the community of researchers and professionals working on AVs. Hence, we wanted to find an area which would make our work interesting to parts of the community.Very early in our first semester, we managed to connect with Melissa Cefkin, an anthropologist working in the development of AVs, who ever since served as a valuable sparring partner. Moreover, we attended a conference on AVs and human factors, and realised that most of the research seemed to focus on only one group of road users: pedestrians. After some contemplation, we found what we were looking for. We decided to produce qualitative insights about cyclists and their interactions with cars and drivers, as Copenhagen provides a unique setting of cycling culture, which we expected to be particularly challenging for a potential deployment of fully autonomous vehicles (fAVs); a challenge we wanted to be the first to reflect on.
 

2. explore & Create insights

So how do you create insights on challenges that lie in the future? This is a typical contradiction for design anthropologists. One can investigate the present and the past, but the future? Our formula was to: look at the past and the present, provide some extra knowledge about what would be changed and envision a future to discuss potential problems. We started by exploring todays’ interactions, needs and challenges of cyclists, then provided them with as much knowledge about AVs as necessary and as easy to understand as possible, before we reflected on potential risks, concerns and opportunities together with them, and finally let themselves build the ideal AV serving their needs as cyclists.
Therefore, we used a set of methods, we were taught in a course running parallel to our project being called: Facilitation of Design Processes and Innovation. We initiated our research with a cultural probe study including 10 cyclists running for about a week. This primarily served to gain inspiration and preliminary insights into the cycling culture of Copenhagen.
 

The outcome were maps documenting challenges on the cyclists’ daily routes, and diaries providing insights into the individual practices, both of which functioned as a great basis for preliminary ethnographic fieldwork. Conducting the right form of ethnographic fieldwork was another challenge. We had many different ideas, reaching from participant observation, recording first-person video material from three different angles on the bike, following bicycle delivery services, to analysing video material from an EU project that tested autonomously driving public buses in different cities. Additionally, we had the possibility to test out some of the methods, during the preliminary fieldwork, which helped us to settle on the most productive approach. In the end, though, it was also a question of time and resources - both being limited, we finally decided to stick to what was most feasible: passive and participant observations. It was not yet the time to go out into the field, though.

Before doing intensive observational work, we wanted to spend more effort on foreshadowing the areas we needed to investigate, and followed up on the probes with individual interviews. To dig even deeper into the specifics of bicycle-vehicle interactions, we conducted a future workshop with the same cyclists that had participated in the probe study. At the workshop, we provided the cyclists with the opportunity to express and negotiate their concerns and needs for a future with autonomous vehicles. Analysing two videos on near contact situations enabled us to reveal problems in the communication between different road users and created a common language between the participants. After a short knowledge input on the status quo of AV-technology, we could finally spark discussions on how communication and interaction would be challenged by autonomous vehicles. The produced knowledge was a necessary basis for the prototyping phase, where our cyclists could explicate and materialise their thoughts in their own version of an fAV, stress-testing it in their own traffic scenario.
 

As can be seen in the pictures, even very low fidelity prototypes might suffice to build a simulation and spark the necessary negotiations to create insights. Eventually, it is not the prototype itself that shows a lot, rather, it is all the discussions emerging that reveal the interesting insights. In our case, the prototype simply served as an object that facilitated discussion and negotiation.

The initial analysis of the cultural probes, preliminary fieldwork and future workshop provided us with a sound amount of questions and concerns that needed (some of which still need) to be further investigated in the field. To document our fieldwork, we made use of elaborated field notes, pictures, video recordings and sketches. While recordings, sketches and notes helped us a lot to understand the data, it is a completely new task to present it. In our last workshop - the co-design workshop - we had two different goals. First, we wanted to present our generated insights, and second, we wanted to facilitate knowledge exchange and ideation. We succeeded in gathering different actors, which we thought could contribute to discuss a potential deployment of fAVs in Copenhagen with, while still focusing on the needs of cyclists. To present our insights, we made use of several methods that allowed us to disperse our input throughout the 5-hours workshop. We invited two of the cyclists from the future workshop who presented their prototypes, we held a PowerPoint presentation, and we printed out personas and data cards (see below).
 

As you can see above, these data cards contained snippets of our raw data, mostly combined with quotes from our cyclists with a sketch or an observation from the fieldwork. The advantage was that they were tangible and initiated discussions between the participants and allowed people to have a closer look for a longer time. Furthermore, it allowed our participants to take the data cards with them, as Melissa Cefkin did for example. To read more about any of our approaches please find our report with an extensive section on our process.

 

3. Don’t underestimate your data...

Designing and executing cultural probes, two workshops and fieldwork in three months is challenging, but we advise students not to be afraid of testing out a variety of methods in a short amount of time. In combination, they can produce rich insights that deepen the understanding of the problem space. After not even three months, we had gathered gigabytes of pictures, dozens of fieldnotes, about 20 hours of video and audio recordings from interviews, two workshops and fieldwork, and more than 100 Post-Its on concerns and ideas from the co-design workshop. Retrospectively, we were impressed by the amount, but being in the process we never really had the feeling that our data was rich in insights - at least until one point. Only when we took one step back from writing our report, printed out as much data as we could, hung it up on walls and put it on tables, and looked at it again, and again, we were able to find topic clusters. Thereby, we combined the data from all our approaches, and finally found the ‘hot topics’ to write about. This process means work and might start off frustrating, but eventually feels rewarding.

4. Make Your insights valuable...

As said before, the goal was to bring our insights to developers and designers of AVs. Until now, we had a written report, we knew that one person had taken our data cards with her to her organisation in Silicon Valley, yet we had not reached our goal. We aimed for a final knowledge object that would be sharp and short enough to be sent out and presented to different stakeholders in the development of AVs. We cut from 130 pages to 8. While we had only planned to electronically send out this summary document, we unexpectedly received an invitation from Melissa Cefkin, our anthropological sparring partner from the Nissan Research Center in Silicon Valley, from Colleen Emmenegger (UCSD Design Lab) and Jane Lappin (Toyota Research Institute) to come to San Francisco for an expert workshop on driverless vehicles and external human-machine interaction. This invitation was what counted. Even if we still had our project defense some days later - there was no greater feeling at this moment than expecting the reality check on our work, participating in one of the worldwide most advanced discussions on the topic we did research on and of course the request to present our findings in front of participants coming from all over the world. With the workshop we reached our goal of bringing insights to the developers and designers of autonomous vehicles.

At the three days conference (AVS 2017) that followed the expert workshop, it was mainly about networking. Due to some informal talks and the participation in specific break out sessions, we got even more opportunities to spread our insights and share the summary document, which then again, might lead to other opportunities, as for example the request to present our work at the Danish Road Directorate (Vejdirektoratet) in Copenhagen or the possibility to make a poster presentation at the TRA 2018 (Transport Research Arena) in Vienna. But don’t be mistaken, there were 49 participants at the workshop with whom our project report and our summary document were shared, plus another 11 persons we shared our summary individually with. From those 60 people that could have potentially read about our research project, we have received the two mentioned opportunities, have been included in an urban planning initiative on the effects of AVs and shared mobility on cities, and have strengthened our engagement in our follow-up project starting in Leeds, an EU research project on AVs called InterACT.

 

5. Experiment, but reflect (our lessons learned)

As mentioned, sometimes you simply have to try out a lot, testing out methods and reaching out to people, and see where it leads you. This has been our philosophy during our whole project, which with the right amount of reflections will also enable you to get most out of your semester project. As a final add-on, we want to share some of the lessons learned from our project.
 

Brainstorm, Align, Reflect

Deciding on a framework for workshops or fieldwork, combined with being time pressured, might lead to more chaotic and stressed group processes. Everyone knows the feeling of being interrupted or not listened to, resulting in ideas being shut down before even discussed. A way we ensured that everyone could have a stake in the decision being made, was to construct rounds of 10-15 minutes individual, silent brainstorming before presenting the ideas, with no interruptions from the group. The listener should take notes and ask clarifying questions after the presentation. No idea is wrong. Once all have presented, the group discusses on which of the elements could be combined for the best concept. This created a great foundation for productive discussions, leading to early alignment and quicker decisions to go on with. However, no decision is 100% final. Staying reflective and changing course is as important, in case you realised that the initial idea was not the best.
 

The 3 T’s of Reflection (Team, Time, Tools)

Designing participatory design workshops for the first time was a bit like shooting arrows in the dark. There is hardly a right answer to what workshop design works or not. Testing your workshop tasks and acting it out as a team might help, but we experienced that you always learn most once you have conducted it in reality. Therefore, when reflecting, you need the whole team, take sufficient time and ideally also use tools to reflect. In our case, feedback questionnaires helped us to improve the ability to facilitate and design workshops. Another tool that kept the group staying reflective were video recordings from the workshop. Going through them focusing on our facilitation, helped to discover the dynamics between facilitators, mistakes or misunderstandings and by that made it easier to give and get feedback.
 

Facilitation is about balance

From one of the feedback questionnaire, we learned that “too many guidelines were given during the design process” while another person said: “it’s sometimes hard with such an open task when you are new to the area”. This is a perfect example that different persons might need different facilitation. Hence, facilitation is about finding a balance. By facilitating too much, you risk disrupting the flow or hindering creativity, whereas facilitating too less might not kick things off as you want it to.

Creating a common knowledge basis might be important before initiating discussions. During the discussions it is a lot about group dynamics. Making sure that everyone is heard and no one is too dominating is the responsibility of a good facilitator.

Aligning the group of participants to discuss the same topic is another task that can go wrong. Especially, experts having different backgrounds, coming from different organisations, might mean that they also have different goals, values and opinions - ultimately meaning different interests during the workshop. Doing the mistake of not aligning these interests to a common goal might block ideation. Therefore, a facilitator might plan extra time on discovering those interests, goals, values and opinions first.

Staying flexible, rather than sticking to every detail of your plan is important. It is difficult to estimate how long an exercise or a discussion would take so always have time-buffers between exercises and don’t be afraid to improvise when needed. It is the facilitator's task to weigh whether it is more important to leave people some extra room for a good lasting discussion, or to stick to the timeline. Sometimes, you even have to go with your gut feeling and throw in a short Q&A session. The balance between staying true to your process and the agenda you presented in the beginning of the workshop vs. your instinct to evoke interesting discussions to get the data you hope for is something you learn by doing.

Make it as clear and simple as possible. Making clear who is the facilitator, who is the note taker, who’s responsible for the recordings etc., is not only important for your team but also for the participants. This creates a safer and more comfortable atmosphere letting your participants know whom they can turn to for questions or remarks. Keeping the structure as simple as possible, is important as discussions usually take longer than you expect.

Finally, support each other. The work of the facilitator is intense, so it is important that the facilitator and co-facilitator support each other. Yet, there needs to be a balance. It is important not to overrule one another. It should be stressed that any differences that should occur should never be discussed in front of the participants, as this could easily create confusion and a bad atmosphere.
 

You get what you build for

Something designers of workshops may never forget, is that all the knowledge produced in a workshop is only the combination of the individual knowledge of all the participants attending the workshop. Recruiting, hence, stays a major point for reflection. Consider well whom you invite and reflect on whom you leave out.

 

More Information

...for more information or if you want to see how some of our methods were conducted, follow the links below!

- 8 PAGE PROJECT SUMMARY

- FULL PROJECT REPORT (incl. a very elaborated process analysis)

Feel free to contact us any time:

Aleksandra Taniberg (Email: atanib16@student.aau.dk)

Markus Rothmüller (Email: mrothm16@student.aau.dk)

Pernille Holm-Rasmussen (Email: prasmu16@student.aau.dk)

Signe-Alexandra Vendelbo-Larsen (Email: svende16@student.aau.dk)