Earlier this year I wrote a highly shared blog post regarding how a vast majority of entry level jobs could be replaced via some sort of M2M solution. Further to that thinking comes this Financial Post article which estimates that almost half (47%) of all current jobs could be phased out using some sort of automation (ranging from actual robotics to software packages).
I think most people can comprehend how some jobs (such as a technician on the line at an automotive plant) could be replaced, as there is a lot of repetition to the work and a robot may offer similar or even better accuracy on this work. However, some of the jobs that were mentioned would not have been on my radar for replacement, such as loan officers and research assistants.
One of the main obstacles to being automated is the ability to gather and process information by being able to predict outcomes based on patterns. For many jobs, there was just too many variables/situations that needed to be accounted for by previous technologies. However, with the rise of Big Data, it is becoming possible to gather more data than ever before, from more sources than ever before.....and most importantly, using analytics in ways that were never available before.
In another previous blog post, I combined two of my favourite passions (M2M and Baseball) to talk about how Billy Beane has eliminated many internal scouting positions using Big Data (just called a fancier name....Moneyball). For now, let's focus on how M2M/Big Data is furthering this trend on a more macro basis.
Better information = more situations to predict
In the case of M2M, it is often in the form of sensors or receiving greater amounts of information from remote sources. This includes things like knowing more information from a remote wind turbine (to be able to better predict maintenance schedules and mechanical issues) to a company-owned vehicle (to better optimize costs and productivity) and vehicle counting solutions (to better optimize traffic flow).
Big Data solutions take this information and combine it with much more. Take the unmanned train that takes tourists all over the strip in Vegas. The first thing that their system must do is take a lot of sensor data (to see what is going on in all parts of the system). However, sensor data alone is not enough to replace the intelligence and experience of a train operator. That information needs to be taken back to some sort of logical centre to be able to make decisions. So, if there is unexpected water on the track, some intelligence is needed to determine the optimal speed, as an example.
More detailed information = more detailed prediction models
One of the biggest issues in using automation to replace humans is the inability for the machine to see "the bigger picture". To use the Moneyball example, all of the data in the world to tell you how good a player may be is useless if they have a drinking problem that forces them to miss many games. This is something that a computer likely would not tell you, but a well-seasoned baseball scout might see. This was always a huge issue for automation...what happens if things do not go perfectly as planned. Again, M2M can help here.
One of the values of a skilled technician is his/her ability to have a good understanding of what is going on with the machines and being able to quickly repair items when they fail. However, with the complexity of many machines, there are so many variables that go on, it is virtually impossible for them to know everything that is going on at a machine in real time. M2M helps here....by providing much more accurate details as to the "moving parts" of a machine, it is easier to predict failures.
But, it goes much deeper than that. With the right information, engineers can run much more detailed models as to the machines....they can better predict how machines will work under all conditions (using extreme amounts of data from live machines) to determine what is needed to better optimize machines. They can also combine this with proactive monitoring tools to change settings (based on actual usage) to better optimize the machine for use....while it is in use. This will never totally eliminate the role of a technician, but it will greatly reduce the demand.
Reducing demand is not that different from elimination of roles in some ways
As the article mentions, most law firms do not feel that computer software will ever eliminate the need for all legal research, only a large portion of it. This means that the role of Legal Researcher will be diminished, which may discourage many young people from entering into the field. With a lack of new talent entering into the field, it will make the "superior" researchers more expensive (as they may be in more demand), which may in turn cause more people to continually look to better the existing software-based research programs. This "circle" means that the job will be on its way to obscurity, even if it does not become obsolete.
Bottom Line
I don't think that robots will be replacing all human roles in the next few decades (there will always be a need for Police Officers). iRobots will be just known as an underrated Will Smith movie and the annoying little vacuum cleaners that terrorizes my cat and eats my iPhone cables. However, it is foolish to think that it won't replace a significant portion of workers today, just like advances in farming eliminated 90% of the jobs in agriculture and the Industrial Revolution eliminated many roles in the manual labour area. These realities reinforce that we need to start to train young people in fields that will have a longer shelf life.
As always, Novotech is ready to assist with your M2M needs. Whether you’re looking to control, track, monitor or back-up, Novotech has the solutions and products you need. View our Line Cards and let us know how we can be of assistance.
Stay connected! Follow Novotech on Twitter (@NovotechM2M) or follow me personally (@LBNovotechM2M). We’re also very active on LinkedIn so follow our company page too.