A Camera on Wheels: Situating the Self-driving Car within a Photographic Teleology

Even before the long-promised arrival of self-driving cars on our roads, vehicles are already increasingly equipped with a range of optical and light-based sensors that are capturing automated images of the built environment toward a post-human photography perhaps most succinctly anticipated by Ed Ruscha’s seminal work Every Building on the Sunset Strip (1966). This paper proposes a media archaeology of autonomous vehicles by engaging with an automotive camera and its machine vision algorithms from within photographic history and discourse, drawing on research pursued at Bosch Research Campus near Stuttgart, Germany, on some of the 'situated knowledges' (Haraway) at work in one of Europe's foremost research centers on automated driving. From the bounding boxes of now classical object detection to contemporary semantic segmentation, the talk probes the technical images used to train self-driving cars and traces some of the aesthetic genealogies that might constitute machine vision’s ‘optical unconscious’ (Benjamin). This includes an analysis of the Daimler Cityscapes machine learning dataset comprising 25'000 semantically labelled photographs of urban street scenes throughout 50 European cities which is used by AI researchers at Bosch to preemptively test the capacity of machinic perception to be deceived in so-called 'universal adversarial perturbations', a form of steganography that alters what the machine 'sees' while remaining invisible to the human eye. The paper concludes by discussing the results of a creative collaboration with the AI team at Bosch to insert a reflexive piece of photographic history into this novel and uniquely machine-readable class of images, closing a loop that situates autonomous vehicles within a teleology of photography.