The self-driving R-Bus, circling random routes, served as a mobile art installation

In August 2023, a special art installation took over the streets of Kalasatama. The R-Bus, guided by algorithms, travelled between four stops, randomly selecting its routes to add variety to each journey.

The R-Bus project was a success on many levels. During the project, the bus ran for a total of 17 days, and attracted more than 1100 visitors.

"This was an impressive number, considering that the bus only had room for six people at a time. The bus was constantly bustling with people queuing at the bus stop waiting to get on," says Laura Beloff, Assistant Professor of Art and Media, who was responsible for the project.

Beloff noticed how people, while riding the bus, quickly began to trust the technology, and how the technological system decided for you.

"The ride was taken as a journey rather than a transfer. The attitude towards tech clearly divided people. The experiment forced people to reflect on their opinion of this kind of transport."

The comments about the project showed a split between for and against this kind of development in traffic.

The urban environment required lengthy permit processes

Laura Beloff is passionate about the concept of intelligence and artificial intelligence, particularly their development in recent years. She was particularly interested in how these things had changed in recent years. Her interest took concrete form in the Smart Bus project.

"I would use the term 'feeble intelligence'. The bus was intelligent, but at the same time it lacked the biological intelligence that animals, for example, have."

Originally, the idea was even more radical, with a plan for the City of Helsinki to have a bus that would go to a totally random direction. Kalasatama was chosen for the project because of Helsinki's 'percent principle' where 1% of the area's building costs are allocated to the arts. This set the conditions for the bus routes.

"The project proved to be complex with roadworks and city construction happening in the area, and the surroundings changed weekly. The permitting processes were a hassle, and the companies and businesses that were attracted to get involved were divided in many ways," says Beloff.

Bus makers and coders were important, as was the operator, which was in short supply. Estonian company Auve.tech was involved in the development of the bus and the software.

"The safety drivers were necessary so that someone could take over the manual control in case of an emergency," says Beloff.

The sound sensors were set up on the north side of Kalasatama, near Sompasauna in an area that is a sort of wasteland. 

"There was an interesting variance of sounds depending on the time of the day or weather. We did tests at night where there were some very interesting sounds of the night", says Guillot Suarez. 

In an urban area, human infrastructure reserves sound frequencies with sounds such as traffic and people walking by. The sounds created by machine learning are based on the empty spaces left over, where the sounds of "new organisms" were created.

"Lower frequencies could have sudden sounds such as low rumble from a construction site that made it harder for animal sounds to adapt and survive. Higher sounds were more available, which is why bird or insect sounds survived and were more common", explains Guillot Suarez.

"Maybe at one point in time there were sounds in the nature at these frequencies too", ponders Beloff.

"The work is a reflection of what the world is like for other organisms. We were not so much thinking about a general aesthetic, but producing a concept. A bit more filtering and quizzing was required towards the end", says Beloff.

"This is a playful way of dealing with a serious issue, but the critique is there below. Why do we dominate and have to decide for other species? In my work I try to create a situation that asks questions and leaves you wondering", she continues.

The R-Bus project team included Laura Beloff, Marko Tandefelt, Petri Ruikka, Jani Hietanen, Calvin Guillot Suarez, Sebastian Schlecht, Andrea Mancianti, John W. Fail, Esther Saraste and Auve.tech & Rolan.