How robots are transforming work in the world's distribution centers

In a post Recently, Henrique wrote about how the digital revolution we are going through has affected and how it could affect the jobs we know today. By the way, this issue of “technological unemployment” is an old discussion. In 350 BC Aristotle already pondered on “what would become of the servants if the lyre played alone?”.

The subject is also frequently studied and discussed among economists. Many argue that technological advances do generate unemployment, but temporary unemployment. The replacement of manual labor by machines in the Industrial Revolution would have led to the elimination of positions at first, but in the end, an increase in productivity prevailed that allowed the generation of more wealth and, consequently, more jobs.

A more recent classic example is illustrated by David Autor in a TED Talk last year on the ATMs in the US banking industry. The numbers are as follows: In 1970, before the introduction of ATMs, the number of bank teller employees was in the order of 250. In 2016, that number was on the order of 500, double the amount 46 years before the introduction of this disruptive technology! What happened was that the introduction of ATMs made it cheaper to open new branches, generating an increase in jobs in general. In addition to more jobs, the nature of work has changed as well. Repetitive, low-value jobs were taken over by machines, and jobs began to require more sophisticated work, involving customer relationship skills, problem solving, and product sales (cards, investments, etc.).

Economists Andrew McAfee and Erik Brynjolfsson, however, found that over the past 15 years Technological advances have resulted in a much faster increase in productivity than the creation of new jobs.

In July last year, McKinsey released the results of a survey that evaluated which jobs will or will not be replaced by robots in the near future. In the study, around 2.000 activities were analyzed, the time spent on these activities was quantified and the technical feasibility of automating them in the US was analyzed. Figure 1 shows the results of the study, suggesting that 45% of the time spent on activities in the US could be automated. The study recognizes that other factors such as implementation cost and government regulations also strongly affect the feasibility of automation or not, however, technical feasibility would be the first conditional factor for these automations to occur.

Figure 1 – Percentage of time spent on activities that could be automated. Source: Mckinsey, adapted by ILOS: Where machines can replace humans – and where they can't (yet)

Figure 2 shows these results for 7 activity categories: people management, application of expertise (decision-making jobs, planning and creative jobs), interactions with stakeholders, “little predictable” physical work, data collection, data processing and “highly predictable” physical work. We see that this last category represents a third of all time spent on activities that could be automated.

Figure 2 – Percentage of time that could be automated by activity category. Source: Mckinsey, adapted by ILOS: Where machines can replace humans – and where they can't (yet)

Much is said about autonomous vehicles, for example, and the threat of reducing driver jobs. However, this type of technology is still being tested in highly controllable environments. Robots are pre-programmed to deal with a finite set of foreseen situations, and, therefore, their implementation in logistical processes of load distribution in the coming years seems unlikely. The delivery last mile through drones it also still has a lot to develop, despite recent advances. That aside from regulatory issues. The McKinsey study estimates that 25% of the time spent on “unpredictable” physical work activities could be automated.

On the other hand, when it comes to jobs that involve a predictably high load of physical work, this percentage was quantified at 78%. In a logistical context, we immediately think of the activities carried out in a distribution center. Incidentally, many of these tasks, especially picking, have already experienced a major increase in the degree of automation in some distribution centers around the world.

In February of last year, Fernando reported the case of DVD rental division and blu-rays from Netflix. With the advent of streaming services and digital media recently, this little-known division of Netflix has lost a lot of demand, but the company has decided not to abandon the segment. However, continuing with the same operating structure was not a viable option. The company invested in automating its 33 distribution centers with the introduction of machines of the type amazing arm, which are 5 times more productive than manual operation. The centers continued to need people, but the number of employees needed to handle the discs dropped from over 100 to 25.

More strikingly, we also have the case of Amazon. In 2012 the company acquired Kiva Systems, a company specialized in the development of automation systems. The reason behind the acquisition was to increase productivity to accelerate its expansion process. The robots started to be used even in 2014, and at the end of that year we know that 15 thousand robots were installed in 10 distribution centers in the USA. In October 2015, this number doubled, reaching the mark of 30 thousand robots in 13 centers (https://ilos.com.br/web/forum-de-davos-e-os-robos-na-logistica/). The company reports that automation has cut operating costs by 20%, which translates to approximately $22 million per distribution center. Deutsche Bank estimates that Amazon could cut an additional $800 million if it extended the process to the other 110 centers. Kiva robots take the racks containing the products needed to compose the order to the picking, cutting the cycle time from 60-75 minutes to 15 minutes.

Video 1 – A day in the life of a Kiva robot

Amazon also works with the robo-stow, one of the largest robotic arms in the world, whose objective is to move large amounts of stock, and with more advanced vision systems, which use images to evaluate the load and automatically upload the information to the system, allowing the unloading of a trailer in 30 minutes instead of hours.

Later this year, DHL will begin the implementation of a pilot project using robots locus from Quiet Logistics. As with Amazon, these robots do not completely replace human labor, but the company claims that it is possible to increase the productivity of a warehouse by 8 times. Robots speed up the location and movement of items of picking. In addition to increased productivity, there is talk of increased accuracy of the operation as well, since manual work is subject to human error.

Other implementations include DB Schenker Logistics, which is automating its picking in Scandinavia with the Swisslog CarryPick system (a mini-vehicle that fills picking similarly to Kiva robots), and the use of technology Auto Store also from Swisslog to automate the process of picking of Texas Instruments distribution center operations in Singapore.

Within the supply chain, this “arrival of robots” for the time being is more palpable in the warehousing area, and many of these technologies have yet to completely replace human labor. However, there is no doubt that productivity gains lead to a lower need for human labor, as repetitive activities can be performed with better quality and faster by robots. Therefore, it is inevitable that certain jobs of this nature will decline as the automation of operations becomes more widespread. On the other hand, there will be a greater demand for jobs related to automation, robotics, programming and maintenance of robots.

References:

Radu Palamariu – How are robots impacting recruitment in Supply Chain and Logistics?https://www.linkedin.com/pulse/how-robots-impacting-recruitment-supply-chain-radu-palamariu-7>

Business Insider – Companies that use robots instead of humans.http://www.businessinsider.com/companies-that-use-robots-instead-of-humans-2016-2/#spreads-robot-lettuce-farmers–harvest-30000-lettuce-heads-every-day-1>

Business Insider – Amazon doubled the number of Kiva robots.http://www.businessinsider.com/amazon-doubled-the-number-of-kiva-robots-2015-10>

Business Insider – Kiva robots save money for Amazon.http://www.businessinsider.com/kiva-robots-save-money-for-amazon-2016-6>

IMC – IoT M2M Council – DHL pilots robotics in life sciences warehousehttp://www.iotm2mcouncil.org/dhllocus>