This article discusses the challenges of automating AFM for biomedical screening applications. Since AFM can discriminate between normal and cancerous cells, it could play an important role as screening tool --- if only in situ AFM imaging on biological specimens could be automated similarly to what is routine in the semiconductor industry. Dujardin et al lay out a path for achieving that. The authors adapt a Dimension FastScan AFM (Bruker, Santa Barbara, CA) with a microwell plate and develop scripts for not only accessing all wells automatically, but also robustly recognizing and targeting cells, followed by acquiring and analyzing data. They even devise an algorithm for recognizing tip contamination and cleaning the tip to avoid cross contamination between wells. It’s an impressive tour de force, taking some of the most challenging AFM experiments, that usually consume countless hours in the life of a graduate student, and transforming them into an unattended process.