In this study, I explore the potential of Sidewalk Autonomous Delivery Robots (SADRs) to alleviate traffic congestion and reduce emissions, with a particular focus on the on-demand food delivery industry. As online food delivery continues to expand, the number of delivery vehicles on urban roads has increased, exacerbating traffic congestion and vehicle emissions. SADRs, characterized by their small size, fully electric operation, and primarily sidewalk-based movement, are emerging as a promising technology to mitigate these issues. However, past research on the traffic and environmental impacts of SADRs within the context of on-demand delivery services remains limited.To address this research gap, I utilized data from Coco Delivery, a SADR company based in Los Angeles. Combining these data with continuous approximation (CA), the EMFAC2021 data, and the eGRID dataset, I estimated the vehicle miles traveled (VMT) and emissions of conventional human-operated delivery vehicles under three different scenarios. I then compared these results with corresponding data from SADRs to evaluate their effectiveness in reducing VMT and emissions under the same delivery demands.The findings indicated that SADRs can eliminate 0.7 to 1.59 VMT per order and reduce various types of emissions by 67% to 99.9% under different scenarios. These results suggest that cities should consider SADRs as an effective tool for reducing road traffic and emissions.