An ethical trajectory planning algorithm for autonomous vehicles

With the rise of artificial intelligence and automation, moral decisions that were formerly preserve humans are being put into hands algorithms. In autonomous driving, a variety such with ethical implications made by algorithms for behaviour trajectory planning. Therefore, here we present an planning algorithm framework aims at fair distribution risk among road users. Our implementation incorporates combination five principles: minimization overall risk, priority worst-off, equal treatment people, responsibility maximum acceptable risk. To best our knowledge, this is first vehicles in line 20 recommendations from European Union Commission expert group general applicability to various traffic situations. We showcase selected scenarios provide empirical analysis principles 2,000 scenarios. The code used research available as open-source software. situations where some injury unavoidable self-driving vehicles, how distributed becomes question. Geisslinger colleagues have developed takes account comply emerging EU regulatory recommendations.