Redundancy is the term given to a part in a system which has been 'doubled-up'. The idea is to reduce the chance of failure for the whole system as much as possible.
In aircraft and crucial infrastructure, critical systems are often triplicated. By duplicating or multiplying the number of critical components or functions, the overall system's reliability improves because back-ups are ready to take-over if parts fail.
Because parts can fail unexpectedly, redundant systems need to happen immediately. In computer systems, parts which are in error can be 'out voted' by others. Or, in a suspension bridge, many cables are the form of redundancy. In both cases all parts would need to fail before the system fails.
If parts are reliable, the likelihood of all component parts failing before they are replaced should be extremely small, making redundancy an important principle for energy supplies.
Failover is the term given to the process of moving to a back-up when a part fails. In a suspension bridge, multiple cables are used to carry a load which just one can handle - so, if one weakens, the others are already there to failover to.
In computer systems, failover is usually an automatic system which switches work to a redundant or standby computer or other hardware component. To maintain a continuous service, this is done automatically with continuous monitoring of a system's component parts. Such systems offer a very high degree of reliability as they enable services to be provided with little or no disruption.