I think the idea is less "additional sensors don't add value in various scenarios" and more "we are 100% certain a vision-only system can perform well on existi

I think the idea is less "additional sensors don't add value in various scenarios" and more "we are 100% certain a vision-only system can perform well on existing road infrastructure; we are not sure about sensor fusion systems".