How do I know when to stop a failing technology project?

Compare the remaining cost, risk, and time with the business value still achievable. Stop or reset the project when the core outcome is no longer valuable, the recovery plan is not credible, or continuing would spend more than the benefit justifies.