Cleaning robots in public spaces from a legal perspective

In times of increasing demographic-related labour shortages, there are high hopes for autonomous service and cleaning robots that can be used in public spaces (e.g. car parks or shopping arcades). Especially for cities that have a large number of public spaces, they could compensate for staff shortages. The use of such autonomous systems raises a number of – in some cases new – legal issues.

Which areas of law are affected by the use of autonomous systems? Do the applicable regulations already create a clear, manageable and realistic legal framework that takes into account both operational requirements and risks when using autonomous robots? Or are autonomous robots – metaphorically speaking – outpacing legal developments and therefore need to be tightened up in the near future?

Legally complex - the public space as an area of application

In any case, the legal situation is challenging. In order to be able to assess it, fundamental questions must first be clarified: Are service or cleaning robots nothing more than machines that can be used by the operator like conventional tools? Or are they motor vehicles that require a licence and, if so, in which application scenarios? Will such systems soon be considered artificial intelligence (AI) within the meaning of the planned EU AI Act, which – depending on the risk classification – would entail complex documentation and monitoring obligations?

Public spaces primarily include streets, pavements and passages that are used intensively by people. Sweeping and cleaning robots operate on “dangerous terrain” in legal and safety terms. This quickly leads to traffic, authorisation and liability law, in which parameters such as permitted driving speeds as well as control algorithms and configurations and their safety play a role.

One thing is certain: there is no clear and unambiguous definition of how autonomous cleaning robots should be categorised. It therefore needs to be clarified whether clear and sensible categorisations to existing legal areas and regulations are possible within the framework of legal interpretations or whether there is a need for legislative action.

Data protection as a key challenge

In addition to the aspects mentioned above, data protection should not be underestimated when it comes to robots in public spaces: if robots have cameras, for example, that perceive their surroundings for orientation or for training AI systems, this can be accompanied by problematic recording of pedestrian data and its processing via servers – possibly in other EU countries. The task of the legal team led by Prof. Dr. Michael Veddern (HdM Stuttgart) is therefore not least to identify and evaluate systems that are sensitive to data protection – in the sense of privacy by design – and, for example, almost completely dispense with data processing or limit it to a minimum in the basic settings.

Examples include machines equipped with lidar sensors, such as the robots used by ZEN-MRI project partner Adlatus GmbH. They use laser technologies to generate abstract 3D models of the environment, which exclude personalisable data. Such systems are ultimately associated with the expectation of greater acceptance of robots on the part of the population, which could prove to be a competitive advantage for providers of autonomous systems.

authors: MV, JH und CL