Single Points of Failure and Innovation

John W Lewis's picture
Chat Date: 
Thu, Aug 03, 2017

Many of the systems that we use have single points of failure which can stop the whole system. But many other systems have redundancies and backups. Let's discuss innovations which introduce or remove single points of failure.

Vulnerable systems

If, for example, you have a large, distributed IT system which relies on a single, central processor, when that processor stops, your whole IT systems can fail.

There are many such systems (processes, products, etc.), in many fields, which fail completely due to a single fault. This makes it important that people using these systems are aware of this vulnerability, and have fallback plans to operate in the case of the system being unavailable.

This happened recently to British Airways, and they had known about this vulnerability since the 1990s, as described in this article "British Airways: a brilliant example of how cost-cutting increases costs".  According to Sir Colin Marshall, Chief Executive of British Airways at the time: "But the thing I fear most is our Information Systems going down."

Redundant systems

However, if you are flying in a multi-engined aircraft, and one engine fails, the aircraft can continue to fly on the other engine(s).
This type of systems continues to operate with one or more faults, although possibly with reduced performance or capability. These systems are much more resilient and can be relied upon with more confidence. Also, if the faults are monitored, precautions can be taken to operate more safely until the fault is repaired. In the case of an aircraft, it can be flown to the nearest airfield.


During #innochat on Twitter at 12noon on Thursday August 3rd, 2017, let's discuss innovations which introduce single points of failure, and innovations which remove single points of failure, based on the following questions:

  1. What innovations do you know that introduced single points of failure?
  2. How are single points of failure introduced and how can we avoid them?
  3. What innovations do you know that removed single points of failure?
  4. How are single points of failure removed and how can we remove more?
  5. What can we learn from the history of innovations introducing and removing single points of failure?


All donations go to help defray hosting, maintenance and domain name costs. Thanks!