Very interesting question! I cannot contribute to the reduction of gaseous substances. However, in the past I have done some research on the fractions of particulate brake emissions (you can find some literature on my profile). For this purpose I compared deceleration profiles of different test cycles (WLTP database) of vehicles with conventional combustion engine with those of mild-hevs, PHEVs and BEVs. In parallel, the particle number concentration was measured using condensation particle counters (CPC) to determine the brake emission reduction potentials. I was able to measure a reduction of over 90% under BEV simulation. Even for mild hybrids with a 48 V starter configuration, emission reductions of more than 50 % can be demonstrated due to a significantly reduced number of brake applications and reduced friction energy per brake application. We are currently conducting investigations under real world conditions (real driving emissions) and will also determine the potential for tire abrasion.

The fact that autonomous vehicles act with foresight and accelerate and decelerate rather defensively will lead to a reduction of brake and tire wear.

For further questions please do not hesitate to contact me!

Kind regards,

David

If self-driving reaches car-as-a-service companies (uber is working on it), you might see the strange scenario of empty cars running around, burning fuel while picking up customers. While this is probably a slight emissions reduction (no driver = 80-90 kg less weight), it will certainly look odd and unecological to people watching it.

Anyway, there aren't many studies because self-driving is still in its infancy and we lack field data. There may be a reducion overall, but I don't believe it will be significant. Those 2-4% mentioned by Zeashan look about right.