This posting apperared on August 25 on Peter de Jager's forum. The man who posted it says that he is not a doom-and-gloom advocate.
* * * * * * *
Many systems and services you take for granted are deeply dependent upon embedded systems technology. For example, oil & gas, petrochemicals, electricity generation, rail transport, air transport, sea transport, medical care, telecommunications, steel making, ore mining and many many more are all dependent upon embedded systems. Essentially the fabric of all countries is based upon these systems. Also, many of these industries supply each other, and therefore the problem is compounded many fold.
The basic problem in software is that date computations occur many times and each has to be modified. The basic problem with embedded systems is that one or two problems may occur, but at a more substantial level, i.e., in the hardware or some other key component of the system. Software repair can be done in house or outsourced to a service provider. While there won't be enough resources for every system, many (hopefully) will be repaired to a level where they are operable.
However, many embedded systems are supplied by one particular supplier and therefore cannot be repaired in-house; therefore, many industries are looking to the supplier of systems to provide fixes or replacements systems. (Remember many of these systems are in niche markets and, hence, only a handfull of suppliers are available.) Moreover, this could mean that one supplier may be asked to provide hundreds of slighly different fixes -- each system being taylored to meet the needs of a particular installation. Also, bear in mind that it takes a considereable period of time to install one fix into a live embedded systems environment. Say nine months, once all safety issues are considered. . . .
Drawing conclusions shows:
1) That there are a limited number of people capable of fixing these system.
2) A supplier will have to repair many systems.
3) There aren't that many supplier around anymore.
4) Will the supplier deliver solutions in time to be commissioned, being in mind this make take 9 months, for example?
Summing up is simple. We won't be able to fix all embedded systems across the globe problems in time. This is mirrored in business, in that not all business systems will be fixed in time. Especially those people who are only starting now! It is logical to suggest that preference be given to those industries that provide core infrastucture, i.e., oil & gas, electricity, communications, medical.