Autonomous Autos: How Human and Car Entangle
Flora Lechner and Cristina-Sorina Stângaciu
Our relationship with cars is parsed through cultural associations and sensorial experiences. As technology for self-driving automobiles advances steadily, so too will the relationship between humans and cars evolve: the sensors that enable an automobile to navigate a physical environment can also interface with the human passengers inside. ‘Synthia’, a creation by Flora Lechner and Cristina-Sorina Stângaciu, is an interactive installation that considers the human-like qualities of these automobile sensors and the kind of relationship they might inspire between humans and our vehicles.
Based in the Netherlands, Lechner is an interdisciplinary practitioner who blends art, design, and performance to explore the power dynamics at play between bodies, objects, and spaces. Cristina-Sorina Stângaciu is a lecturer and research engineer at the Computer and Information Technology Department of Politehnica University of Timișoara. Her areas of expertise span embedded and real-time hardware-software systems, the Internet of Things, and power management in embedded devices.
In this interview, they discuss the boundaries and potential of the technology currently being developed in the realm of autonomous vehicles and embedded systems, give insight into the evolution of their project, and reflect on the impact of their collaboration.
Your project ‘Synthia’ blurs the boundaries between humans and cars. How did you arrive at this theme?
Flora Lechner: We were both interested, in different ways, in the relation between body and machine. I stumbled on a very weird article that used car maintenance as a metaphor for human fitness. It showed how to treat your body by explaining how to treat your car, because apparently some people better understand how to treat a car than a human body. It was ridiculous, but very inspiring. This started a thought process of switching around the car and the body. Digging deeper, I found these old-fashioned drawings from Fritz Kahn, which visualise the human body as machinery in a production industry. The sensory experience of riding in a car does link the body and machine—when a car suddenly stops or accelerates, you feel it in your stomach. It becomes difficult to understand where the body stops and the car begins. This sensory experience of the machine brought us to consider the sensory dimensions of Cristina’s work with embedded systems.
What are embedded systems?
Cristina-Sorina Stângaciu: Embedded systems are microprocessor-based hardware-software systems designed to perform particular tasks in the world. They enable autonomous vehicles, for example, to sense and react to their environment. These systems gather data from the environment like we do through sensors, and then use actuators to alter their environment, much like we use our bodies. In autonomous vehicles, certain human senses are replicated: sight with cameras, touch with pressure sensors, and hearing with microphones. Other sensory mechanisms—such as radar, ultrasound, and infrared—have no human analogue.
The car has often been compared to the human body, and vice versa. But as cars get smarter, the mind emerges as a new layer in the human-car metaphor. Your project makes this explicit through the creation of the ‘Synthia’ persona. What led you in this direction?
FL: With self-driving cars, there is a lot of anxiety surrounding the loss of control. By deliberately anthropomorphising the car and endowing it with a human-like personality, we were curious if this would calm some of these anxieties. We are very interested in the question of who is controlling who in this relationship.
Anthropomorphizing could be viewed as dishonest because it masks the technical complexity of the system. How would you respond to this?
FL: I don’t personally see anthropomorphism as a bad thing. Many belief systems grant human-like qualities to objects, but these are often devalued by Western rationalist culture. I don’t find it so bad to anthropomorphize an object in order to make it more familiar. I sometimes speak with my objects and thank them that they are still functioning. I think that there’s a form of anthropomorphism that doesn’t attempt to rank things in relation to the human, but rather to better facilitate communication and understanding.
Autonomous vehicle networks are notoriously complex and difficult to understand. Given the extent that these vehicles are entangled within broader systems of sensing and information processing, how much ‘autonomy’ do autonomous vehicles really have?
FL: Autonomous vehicles are much more entangled with their environment than traditional vehicles. In some cases, the cars can be easily tricked into stopping by simply placing a traffic cone on the street (activists have exploited this behaviour to boycott self-driving cars). This shows that the cars are the opposite of autonomous with respect to their environment and the technical networks they operate within. In another sense, autonomous vehicles free the driver from the responsibility of driving, turning the car into a place for socialising or leisure. But is the driver really free from responsibility? If the car has an accident, who is responsible? The driver, or the person that programmed that car?
CS: I think it helps to think about different levels of autonomy. There is no clear binary between a fully self-driving car and a fully manual car. Even before automated systems, protocols such as speed limits, laws, and road signs all restrict the driver’s freedom. In modern vehicles, there are different levels of automation: automatic braking systems, lane assist systems, or cruise control systems. And the number of automated functionalities is increasing. The question becomes: what is the threshold beyond which the car becomes autonomous? At what point on this spectrum do people begin to feel uncomfortable? I think the thing that makes people uncomfortable is the decisions made by an automated system. That is where we feel our liberty can be compromised.
To follow, I wonder what is the threshold beyond which a simple trigger response is considered a ‘decision’? It seems that this threshold might relate to processes of reasoning.
CS: There are lots of debates surrounding situations when the vehicle needs to choose which person to save in an accident, for example. Does the algorithm save the passengers, the pedestrian, or the driver? There are differences in the way humans and algorithms react in this aspect.
Flora, your work usually involves the creation of handmade and one-of-a-kind pieces. This is the opposite of the industrial methodologies pioneered in the automobile industry. Did you modify your process in response to the themes and demands of the project?
FL: In my projects I usually do every step by myself, but the scale of this project forced me to get out of that loop and to outsource. This was super interesting, and brought me now to a whole new way of working. My process loosely mirrored the actual process of car design. I started with sketches, then cardboard models, then digital 3D models. These models were used to create templates that were sent to a metal fabrication company to be cut. I then assembled the pieces by hand, using 900 screws. In a way, it was a back-and-forth between craft and industrial techniques. I used the computer, but just as a simple tool for drawing: an extension of my hand. I didn’t exploit the formal possibilities of 3D modelling software, but just used it to scale up my 2D drawings and transfer files to the company to cut. In this way, I think I still stayed true to the idea of handcraft despite the more technical workflow.
Large-scale infrastructures like autonomous vehicle systems are often characterised as coherent and legible systems. But in reality, the messy and provisional nature of a collage seems like a more apt metaphor. How do you view the final outcome?
CS: Seeing the work of the research department through different eyes made me notice things I was not aware of before. Seeing through the eyes of the artist or visitor shows that our work, by itself, can be beautiful. I was particularly surprised by the impact that our installation had on children. In the special programmes that invited students to the exhibition, the children interacted very well with our installation. It is important for us to evaluate our work not only in objective terms: the number of citations, the impact factor of an article, and so on. And in this experience, we have seen the beauty of our work, not only the numerical part.
FL: I see the work as a type of collage, because it reflects both on our collaboration, and the diverse streams of knowledge involved. It’s a big puzzle piece with so many different layers, a storage space for all the different interactive elements and sensors that Cristina and her team put together. Some of these sensors existed before, and others were made for the project. The final output also serves as a memorial, inspired by pop-cultural representations of the car throughout history, but also reflecting on the future.
Text by Connor Cock