Auto­ma­ted dri­ving and the lia­bi­li­ty of the manufacturer

Gui­de­lines for auto­ma­ted dri­ving: con­se­quen­ces for manufacturers?

On 20 June 2017, the Ethics Com­mit­tee, led by for­mer Supre­me Fede­ral Con­sti­tu­tio­nal Court judge Udo Di Fabio, sub­mit­ted its report on auto­ma­ted dri­ving. During the pre­ce­ding nine months, the Com­mit­tee had deve­lo­ped some initi­al gui­de­lines for auto­ma­ted dri­ving. Cen­tral to the report are 20 pro­po­si­ti­ons, which deal with issues inclu­ding data sove­reig­n­ty, unavo­ida­ble situa­tions and fun­da­men­tal goals for the domain of auto­ma­ted dri­ving. Right at the begin­ning, in the first pro­po­si­ti­on, the cen­tral idea that auto­ma­ted dri­ving should fol­low is made clear: “Part­ly and ful­ly auto­ma­ted traf­fic sys­tems ser­ve pri­ma­ri­ly to impro­ve the safe­ty of all tho­se par­ti­ci­pa­ting in road traffic.“ 

This is fol­lo­wed by the second pro­po­si­ti­on, which goes on to shape that idea in such a way that the objec­ti­ve should be a reduc­tion in the amount of inju­ry and dama­ge, the ulti­ma­te aim being the com­ple­te avo­id­ance of acci­dents. This is made more con­cre­te in the second sen­tence of the second pro­po­si­ti­on, which sta­tes that the licen­sing of auto­ma­ted sys­tems is only tenable “if in com­pa­ri­son with human dri­ving per­for­mance it pro­mi­ses, at the least, a reduc­tion in the amount of inju­ry and dama­ge in the sen­se of a posi­ti­ve risk balan­ce”. The Ethics Com­mit­tee has thus made it unequi­vo­cal­ly clear what demands should be made on manu­fac­tu­r­ers if they want to licen­se vehic­les with auto­ma­tic or auto­no­mous dri­ving sys­tems: the auto­ma­ted sys­tems must sur­pass the avera­ge dri­ver in respect of dri­ving safe­ty and acci­dent avo­id­ance. It is true that this expec­ta­ti­on is a logi­cal exten­si­on of ori­en­ta­ti­on toward a bene­fit in terms of safe­ty, but it does ent­ail a cen­tral problem.

The­re is on the one hand some empha­sis to the effect that the auto­ma­ted and con­nec­ted tech­no­lo­gy must be desi­gned in such a way that threa­tening situa­tions can­not deve­lop in the first place; expli­cit refe­rence is made to the dilem­ma situa­ti­on, in which the auto­ma­ted vehic­le has to deci­de which of two ‘non-tradable evils’ (los­ses of human life) it is to choo­se; wher­eby in other trade-off situa­tions, accor­ding to the report, the safe­guar­ding of human life always takes prio­ri­ty over dama­ge to pro­per­ty and inju­ry to ani­mals. On the other, the report fails to offer a solu­ti­on for tho­se very cases in which the auto­no­mous or auto­ma­tic sys­tem real­ly is con­fron­ted with a situa­ti­on of that kind. The aut­hors con­tent them­sel­ves with poin­ting out that in a situa­ti­on in which ine­sti­ma­ble objects of legal pro­tec­tion, i.e. lives, are in mutu­al oppo­si­ti­on, the decis­i­on ‘life vs. life’ can­not be pro­grammed in such a way as to be free of doubt, and that the decis­i­on the­r­e­fo­re con­ti­nues to be incum­bent on the ‘moral­ly dis­cer­ning’ and respon­si­ble dri­ver. The­re is no gene­ral impu­ni­ty as regards action taken in dilem­ma situa­tions; the cul­pa­bi­li­ty of the actions of the indi­vi­du­al invol­ved must be exami­ned from case to case. Yet auto­ma­ted action is not cover­ed by any such exemp­ti­on. So the imple­men­ta­ti­on of auto­no­mous vehic­le sys­tems is likely to depend on whe­ther or not the sys­tem suc­ceeds in almost cer­tain­ly pre­clu­ding situa­tions of that kind.

We are, fur­ther­mo­re, remin­ded in pro­po­si­ti­on ele­ven that the basic prin­ci­ples of pro­duct lia­bi­li­ty app­ly ‘nor­mal­ly’ to auto­ma­tic dri­ving sys­tems. This hint to the manu­fac­tu­r­ers car­ri­es the foot­no­te that they are under obli­ga­ti­on to opti­mi­se their pro­ducts and obser­ve the mar­ket on a con­ti­nuous basis. With regard to self-determination in respect of the data gene­ra­ted, pro­po­si­ti­on fif­teen explains that the regis­tered vehic­le kee­per or the vehic­le user is to deci­de about whe­ther or not his or her vehic­le data may be pas­sed on and if so how they are to be used. Howe­ver, this free will aspect in respect of the pas­sing on of data should be prac­ti­ca­ble for the user of the data too. 

In con­clu­si­on, it should be noted that whilst the report of the Ethics Com­mit­tee does not deve­lop a direct­ly bin­ding effect, it could, indi­rect­ly, con­sti­tu­te an initi­al indi­ca­tor for the legis­la­tors. That said, it does not offer an approach to sol­ving the pro­blem of the dilem­ma situation.


Stay up-to-date

We use your email address exclusively for sending our newsletter. You have the right to revoke your consent at any time with effect for the future. For further information, please refer to our privacy policy.