I'm trying to figure out an efficient way of snapping lights to my headlight geometry on an exported alembic. The alembic is being exported outside of Houdini. I have 18 pieces of geometry that need lights attached and one way of doing this would involve creating a point group at the SOP level for each one and creating a corresponding rivet object that will parent the lights. Since the rivet object takes the average of each point in the point group I have to create 18 solo point groups ( as far as I understand ).

Is there a better way of doing this?

If I export my alembic with the vehicle at the origin on a frame before the animation starts perhaps I can place all of my lights once and have just one point/rivet combination. I'm trying to avoid this workflow as adding this extra frame will become a bit of a nuisance as the animation is changing frequently.

You could link the lights through a direct reference to a point inside a SOP. Use an expression, like this, inside the transform fields of the light.

The X axis. (63 is an arbitrary point number available at the OUTput)
The Y axis.
The Z axis.