Emerging concepts for on-chip biotechnologies aim to replace microfluidic flow by active, molecular-motor driven transport of cytoskeletal filaments, including applications in bio-simulation, biocomputation, diagnostics, and drug screening. Many of these applications require reliable detection, with minimal data acquisition, of filaments at many, local checkpoints in a device consisting of a potentially complex network of channels that guide filament motion. Here we develop such a detection system using actomyosin motility. Detection points consist of pairs of gold lines running perpendicular to nanochannels that guide motion of fluorescent actin filaments. Fluorescence interference contrast (FLIC) is used to locally enhance the signal at the gold lines. A cross-correlation method is used to suppress errors, allowing reliable detection of single or multiple filaments. Optimal device design parameters are discussed. The results open for automatic read-out of filament count and velocity in high-throughput motility assays, helping establish the viability of active, motor-driven on-chip applications.