Robotics is such a broad topic that the decision on whether it's life-changing or not is actually completely relative to what sort of robotics we're talking about.

Automation is already in progress. Thousands of factories involve automated assembly procedures that are, in true essence, robotics. There are bomb-squad bots that venture into hazardous areas, undersea robots that can travel to the depths without need of expensive and spacious crew compartments, and lets not forget space probes. Even the Voyager probes were in their own respect robots. The modern Mars landers are more obviously robots because they are equipped with a more decisive IF-THEN engine to allow them to take appropriate samples.

Perhaps what we're venturing into is the area of self-contained robots, once that require little to no user intervention. The aforementioned Mars landers were capable of performing their duties without human instruction, since the eight minute radio gap would make remote control pretty unwieldy.
Or maybe robots with self-sentience. Sentience is the state of being self-aware and conscious of consequence, capable of foresight. Foresight is the defining barrier between Humans and lesser animals, as we can look at a situation and from it calculate a response beyond what could be determined only by the immediate sensory information.

For many years I've strived to create a ficticious world that is the image I see as our future. In it, there are such things as self-aware AI but also non-sentient automatons such as robot-controlled cargo lifts, trains, and factories. The AI are equipped with a very complicated IF-THEN engine that rivals the workings of a Human mind. These entities are not entirely free-thinking, since they have a five-step directive core that limits their activities, two steps based on Asimov's laws of robotics. These steps are:
1. Entity must follow orders provided by Supervisor.

2. Entity cannot cause the death of any Human

3. Entity will preserve self, so long as it does not conflict with Directive 2

4. Entity must answer all questions truthfully to the best of their ability.

5. Entity may only act according to Directives or Current orders

With these particular rules in place, I don't see how a robot Apocalypse could be possible. The most notable are number 1 and 5. These directives prevent an entity from operating outside of orders, and they need to obey every order given. This prevents several of our popular pop culture apocalypses.
Asimov's robots imprisoned humans to save them from themselves. However, the robots were never issued a command to imprison humans. An act of this sort violates directive 6. Unless an order was given. But that's not the robot's fault, it's their supervisors. In the U.S. legal court, a bullet isn't blamed for the death of its victim, the gun's wielder is.
The Terminator-style apocalypse: Terminator violates so many directives it's almost amusing. Firstly, it plays off of Cold War tension. In the third rendition, the future is said to go as follows: Skynet nukes Russia, which nukes the United States in turn. Politically, this would be unlikely as Russia no longer has the capability of firing half of their nuclear ordinance. Most of their submarine fleet is rusting in harbors, unmanned, with no one around to turn a launch key. Secondly, Skynet's bombing violates directive 1, 2, and 5. It was not issued an order to fire, it is causing known harm to Humans, and it is not allowed to generate this sort of objective.
The Matrix is perhaps the more plausible. The first fault in Matrix logic is that the X-amount of BTUs and amps created by Human bodies are part of a self-sustaining loop. That is, we are not loose power plants running about, there isn't a lot of surplus energy in our bodies we are not using, and in order to generate sufficient net power surplus you'd need to run a car you'd need several hundred humans. Secondly, the robots in the Matrix were not programmed with directives that involved Humans. According to the Animatrix premise, the robot response was a reaction to Humanity freaking out and trying to bomb 01 off the face of the earth.

So by these rules robot apocalyspes are impossible without factoring in human fallibility. By these five rules a supervisor could order some severe damage unless the robot hangs up on Directive 2.

We should also look at the ramifications of automation. Obviously, no action in the world is without consequence. The decision to drive to work generates polluting fumes. The decision to wear a set of pants puts more wear on them. Intensive automation would eliminate thousands of low-skill labor jobs such as train engineers, factory workers. As sad as it is to admit it, there are thousands of people who are unable to perform above these low-skill jobs.
But, at the same time, it creates machines that are able to function in hazardous situations, such as areas with high radiation, toxic fumes, and explosive hazards that would threaten human operators.