In plasticulture production, smart spray technology can detect weeds and apply herbicides only where needed in the area between raised, plastic-covered beds (row middle). This technology has the potential to reduce herbicide use and lower input costs. A prototype smart spray system was developed at the Gulf Coast Research and Education Center (GCREC) in Wimauma, FL, that utilizes YOLO-V3 convolutional neural networks to differentiate broadleaf, grass, and nutsedge weeds in row middles. Two sets of field experiments were conducted to determine the efficacy of smart spray technology using a combination of preemergence (PRE) and postemergence (POST) herbicides. All treatments reduced weed density, and targeted applications were as effective as banded treatments. Overall, including a PRE herbicide tended to lower weed density compared to POST applications alone, regardless of application technique. Two banded PRE herbicide applications and two targeted POST applications reduced herbicide use by 52% and 13% compared to banded PRE and POST applications in Experiments 1 and 2, respectively. The reduction from two to one PRE-herbicide applications did not result in an overall herbicide use or cost reduction in Experiment 1, as the decrease in PRE herbicides resulted in increased POST-herbicide usage. In the absence of a banded PRE application, targeted compared to banded POST applications, herbicide usage was reduced by 40 to 67% in Experiment 1 and 79 to 84% in Experiment 2. Smart spray technology is an effective weed management tool for row middles in plasticulture production systems with or without PRE-herbicide applications.