If we use multiple satellites traveling ahead of Earth on its orbit we improve detections of sunward objects over satellites in NEO-Surveyor-like orbits by tens of percent and enables seamless, automatic follow-ups.

A University of Delft research paper indicates we should have over 20 space telescopes in early warning networks.

Enabling Interceptors and Diverters: From Detection to Deflection Via SpaceX Starship

Early warning would then be used to triggers deflect-on-demand via pre-staged assets. With 5+ years' notice:

SpaceX Starship can launch a fleets of 10-20 ton Hammerhead probes that would be DART successors. They woul have with 500kg impactors for velocity boosts (6-10 km/s Δv on 25m targets). A single Starship could deploy 5-10 and they could intercept with 1-year flight times.

We could use nuclear diverters for some larger objects. For 4-25meter clusters we should developed enhanced Orion-style devices (yield 0.1-10 kt) for standoff ablation, launched reactively. Starship's volume fits 4-6 units per flight.

It will also be critical that the detection satellites reliably detect objects 25–50 meters in diameter (and ideally even smaller to 1-25 meter sizes). This is demanding high sensitivity and increased longevity. With improved resolution and constant space-based observation, humanity will have ample warning time for even Chelyabinsk-like impactors. This is proposed by Joe DeMartini.

A constellation of infrared telescopes optimized for SpaceX Starship deployment could be used for a comprehensive network of telescopes to detect all threatening asteroids 4 meters or larger. A multi-node IR (infrared) observatory network could designed to close blindspot gaps and give years of time to enable proactive deflection.

ESA's planned NEOMIR mission will be located between Earth and the Sun and will act as an early warning system for asteroids 20 metres and larger that cannot be seen from the ground.