Can analog sensors replace limit switches?

I recently read an article that said that traditional overpressure monitoring using on/off limit switches is being replaced in many cases by more sophisticated analog pressure sensors , signal conditioning circuits, and algorithms that can more fully assess the sensor and overall system condition than a “simple” switch. The article made a good comment on the limitations of limit switches: they are susceptible to drift and aging, are difficult or impossible to recalibrate once installed, are difficult to test in-circuit, and do not tell you anything before a problem or limit condition occurs. Analog sensors and associated circuits have come down in price and become more common. They can provide more information and tell you what is going on because they report the pressure data before the limit condition occurs. They can tell you the pressure in real time and to some extent their own condition. One can also test, recalibrate, check, etc. these sensors and circuits even if they drift or age. I agree with all of this, but not all of it. Yes, getting continuous readings from a pressure sensor and looking at trend curves and rates of change is better than the “oops, we hit the limit” information that a simple on/off switch can provide, no doubt about that. But to achieve this extra intelligence requires putting many more parts in the safety loop (signal conditioning, processors, memory, and annoying software). As we all know, the more parts in a design, the more potential for latent or subtle design flaws or even outright failures. The final conclusion: more complex systems may be smarter, but they are also more prone to more problems that may not be visible until they occur. Many years ago, I worked on electronics used with hydraulic pumps that moved many gallons of fluid per minute at pressures up to thousands of psi. I was relieved to learn that the pump had a simple safety loop consisting of a pressure limit switch (several, in fact) that was wired directly to the pump's motor control relay. Yes, this pump failed several times, so there is a deep reason why it is still used today, thanks to its ability to work in parallel with more advanced electronics, and in fact it can operate independently of the electronics. Here are some specific examples and thoughts shared by two engineer friends that further support the above points: Bearchow: "I personally have two impressive electronic systems with sophisticated microprocessor control circuits, a Carrier air conditioner in a recreational vehicle and a Seffler very large capacity water pump. Both systems often either stopped working or had a fault light on, and the instructions in the manual were to send them for repair (at the time I was sure that all mechanical parts were good). Fortunately, both systems allowed me to disconnect the processor part and build good old-fashioned relay and switch control circuits (for example, the freeze limit switch on the AC condenser was originally electronic, but now uses a mechanical switch from a 1975 Chevy "Impala"). On the water pump, they fortunately left the motor wires and pressure switch (yes, at least the pressure cutoff is a real switch), so I used a relay between the two because the switch used with the microprocessor had uncertain current carrying capacity. Both systems have worked reliably for several years since I eliminated the microprocessor." Stucbrown: "The KISS principle (Editor's note: KISS stands for keep it simple and reliable . "Stupid" should always be the watchword of engineering. Just because something can be done in a new and different way doesn't mean it should be done. More parts and software don't necessarily improve MTBF. If young engineers look at me strangely, I tell them I'm used to using tubes. That always gets a laugh, but I'm not kidding."