Where to put the cameras

A city wants to place CCTV at intersections so that every street is monitored from at least one end. Each camera costs the same; minimize the count. That is vertex cover on the street graph. The same problem appears in IoT sensor placement (cover every link in a mesh network), in network monitoring (place taps so every fiber is observed), in chemistry (place an inhibitor on every reaction edge), and in patrol scheduling (assign an officer to every corridor). The 2-approximation is good enough for almost every commercial deployment — you rarely care whether you used 51 cameras instead of the optimal 47, especially when "optimal" requires an overnight MIP run.