The original plan for Hadrian’s Wall featured no forts on the Wall itself, only milecastles and turrets. Troops stationed at the milecastles and turrets would have used semaphore and lighted flares to communicate with forts like Vindolanda that lay along the Stanegate road to the south. Construction was already underway when it was decided that a new series of forts should be built on the line of the frontier Wall.
The archaeologist David Woolliscroft has researched the evidence for the use of signalling systems by the Roman army using Hadrian’s Wall as a case study. He mapped out all of the Roman military installations (turrets, signal stations, milecastles and forts) and checked their inter-visibility, i.e. whether you can see one site from another and can signal between them. He found that virtually every single milecastle and turret had a direct view to a fort, or known relay tower, which allowed for a very efficient signalling system. He has also made a convincing argument that the line of the Wall and locations of installations were at least partly determined by the need for signalling, sometimes even at the cost of tactical considerations. Other Roman frontiers have since been shown to have worked in the same way. You can find out more about the Roman army and the construction of Hadrian’s Wall at the Roman Army Museum near Greenhead.