This means that, assuming a single-agent projection area of a single pixel , any agent moving slower than 41µm/s could be imaged at the highest possible resolution (latency of 0.725 s) whilst maintaining accurate light projection in relation to their position. More typically, the DOME is operated at a resolution of 1920 × 1088 pixels for which the latency is 0.25 s. At this resolution, agents moving slower than 120 µm/s can be accurately tracked by projected light. For faster moving agents, lower projector resolutions can be employed, larger light projection areas used, or both. More advanced algorithms can also be used to predict the location of an agent in a subsequent frame based on its current trajectory, however, this has not been necessary for the applications we have explored.
Implementing building blocks for collective behaviours
The ability of the DOME to interact with many microscale agents in parallel and in a closed-loop manner makes it a useful tool for the engineering of microswarms. To demonstrate this, as well as the functionality of the system more broadly, we implemented three essential building blocks that pave the way for future interactive collective behaviour. These building blocks, namely signalling, stigmergy and motion control, were implemented using the algae Volvox as our microagent. Volvox was chosen due to its innate capability to move and sense light \cite{Drescher_2010}, and because the spherical shape and size of its colonies (350-500 μm in diameter) are easily visualised under a microscope (Figure \ref{115123}).