I'm interested in knowing, how exactly would you suggest institutions go about fixing the replicability crisis? With how things are as of now, I can't think of

I'm interested in knowing, how exactly would you suggest institutions go about fixing the replicability crisis? With how things are as of now, I can't think of any way the scientific community can easily fix this problem.