A Blog by Jonathan Low

 

Feb 10, 2017

The Reason Robots Aren't Replacing Humans In Key Areas

The financial and operational cost of automation favors humans in numerous tasks. JL

Ron Harbour and Steve Scemama report in Forbes:

One European auto plant that invested 10 million euros in technology that would install windshields on cars, replacing people. Maintaining such technology required twice as many workers as the company had employed installing the windshields in the first place. The most automated plants too often fall into the bottom quartile of plants based on productivity. People are the most flexible form of automation. They can do anything. You just need to train them.
For workers, it’s intimidating to hear of industrial digitization plans that envision handing over anywhere from 60 to 80 percent of processes to robots and other programmable machines in the not too distant future. But while there are certainly highly repetitive jobs bots would perform more efficiently and economically, automating alone is not always the best path to higher productivity.
Smart organizations learn quickly enough that if they place efficiency above a smooth organizational transformation, they may find their automation efforts fail to improve their companies’ performance. The real key to developing a competitive edge in an age of evermore automation is striking the right balance between people and robots, and evidence abounds that it’s not necessarily the most automated factories or service organizations that rise to the top.
The automotive industry, among the first to embrace robots in the manufacturing process, provides a working example of why companies cannot simply replace employees or fail to retain and retrain. Stark productivity differences exist between the industry leaders and laggards, in large part based on the efficacy of their automation efforts. One result: Some automakers require as much as six months to transition to producing a new vehicle, while others need no more than a day.
Robots Versus Humans
At the root of the discrepancy is an appreciation of which jobs robots do more efficiently and which require a human touch. Leading car companies have almost completely automated their paint and body shops. These are jobs that require constant repetition and consistent quality and often present safety and ergonomic challenges. Although lead-based paints aren’t used anymore, working in these areas still could expose workers to a bevy of unhealthy chemicals, making these the quintessential kinds of jobs robots have been designed to handle.
On the other hand, assembly lines — which must deal with the multitude of options on new models from side airbags to built-in vacuum cleaners — continue to heavily rely on a human workforce. To handle today’s highly customized vehicles, with as many as 55,000 parts for the variety of electronics and other bells and whistles offered on autos, requires the flexibility of human workers who can adjust to changing needs and innovations without extensive reprogramming. It’s also important to understand all the costs involved in automating. Take, for instance, one European auto plant that invested 10 million euros in technology that would install windshields on cars on the assembly line, replacing the people who once did the job. Admittedly, the new machine was more consistent in applying the adhesive to hold the windshields in place, but it turned out that maintaining such highly sophisticated technology actually required twice as many workers as the company had employed installing the windshields in the first place. In the end, the most automated plants too often fall into the bottom quartile of plants based on productivity.
The Customization Boon
If you look at the most agile, most cost effective and highest quality operations, you will notice that head counts haven’t plummeted over the past two decades, although workers may not be doing the same jobs. A good example is a North American production facility where a U.S. automaker is producing one of the newest and most customized models in its fleet. Automation and robots have been embraced for decades at this plant: Between 2005 and 2015, the company increased the number of bots in the paint and body shop alone to more than 1,000. Yet, the number of plant workers has only declined about 8% during the same 10 years — even as production in the last couple of years fell by almost 100,000 units because of the changeover to a new model.
What kept so many humans on the job? The key was the high level of customization in the latest model and savvy recognition by the manufacturer that keeping robots and automation reprogrammed to meet constantly changing needs may have delayed the transition and, short-term at least, made it more expensive.
Drawing from the experiences of automakers, change needs to be evolutionary, even if the impact of automation is ultimately revolutionary. When introducing bots and automating processes, managers should look to solve specific problems by implementing low-cost solutions, not automating large swaths of functions all at once simply in the name of efficiency.
Who's Critical?
To bring along employees, managers must introduce automation in steps. If they go too far too fast, they risk losing critical know-how as employees jump ship or are pushed off. A priority must be identifying and retaining the employees critical to re-engineering processes down the road — as well as those people needed to ensure the effective management of the bots and automation just incorporated into the workflow.
This revolution promises huge changes as physical infrastructure transforms; offshore capabilities are repatriated; more services become self-service and virtual; and customers begin to link more with robots. Companies will need to give employees new roles and responsibilities, training, and even new career paths as many transition into a new breed of professional with both business and technology skills who can manage both bots and humans in the future.
Along the way, managers should always bear in mind the lesson automakers have already learned: People are the most flexible form of automation. They can do anything. You just need to train them.

0 comments:

Post a Comment