A Case From Russia: Optimize Your Warehouses With The Help Of The Fitness Bracelets And Neural Networks

Today high technologies are in everyday life longer and harder than developed. If once it was hard to trust the electronic device is car navigation, but today its hard to believe a car can go without the driver. Same thing with new production technologies — Amazon ready to ship from stock-airship consignments drones, and most of the warehouse is still associated with dirty and poorly paid manual work. Technology, which can radically change the workflow itself, and out of it today available to every company. On the market of warehouse automation and system integration we have reached in 2004, in the midst of development of Russian market of foreign vendors.

Every second of them was a “leading European” cost solutions exceeded $200 thousand, and the functionality was so limited that when dealing with them mind unwittingly came Association about Indians and beads. Even had to hear “forget about the development and sell good European software”. To be able to compete with companies whose turnover in that time, hundreds and thousands of times was more of our own, we decided to bet on innovation. Realizing that the pool of opportunities for automation of warehouses simple means are already exhausted, we were looking for something completely new — something that will allow you to “run faster”, it would seem that not having the capacity to do so.

And it came by itself — many employees in our company, play sports and know about physical activity and how it is important to competently approach to training, and therefore any physical labor. We realized that it is necessary to change the approach, the work culture, to go from a man of his capacity and capability. By 2006, we had accumulated a voluminous theoretical base and power equipment it is allowed to use technology on the basis of the emulator neural networks. The market was absolutely not ready even to use the term in this context, at industry conferences looked at us as warlocks, selling magic powders, and once at the end of the report there was a loud and long argument with shouts and waved fists. Only one person in the front row quietly listened to all the arguments, reading the newspaper, then folded it and Vesko noticed.

“The Japanese about that 10 years ago said.”. We invited one of its customers to launch a pilot project analysing quality of shipments, which would form part of teams of employees on the grounds of quality and performance, and to identify potentially at fault problems. And we believe. Looking ahead, it was worth it — the forecast accuracy was more than 80%. The complexity of the client was high rotation of staff, but staff often returned to the warehouse “to make money”.

That is, collecting data during the employees work, you can use them later, when the employee will return to work. Another issue was pronounced personal relationships between employees when one team was literally impossible to put certain people together was fraught with conflict, or, conversely, long intimate conversations that always resulted in low quality of shipments. The task was to analyze the possibility of replacing the omniscient foreman information system (if not fully at least partly). The area of the warehouse of the customer amounted to approximately 7 thousand square metres, were in change of 30 employees.

Implementation time was 4 months, the measurements of parameters were carried out monthly. In the result of this work the data we have analyzed, compiled and presented at the industry conference in October 2006. Today we smile and remember that time, because these tasks much more efficiently you can solve a completely different and far more simple methods, but often use of the potential of neural networks required is not so much to solve the problem how to find solutions. We realized that the tool is working, and it can be used. In 2009-2010, we used the functionality of neural networks in order to identify factors that must be considered when calculating the parameters of trade flows.

Results we were able to use to develop their own methodology of technological design. At that time we made one of their biggest mistakes — that was our marketing fail. Actually having the finished product and spending to develop about 7 million rubles, we paid attention to its functionality and algorithms, but not the market. As a result, graduating with a passionate resolution of technical issues, we found continuous losses and the complete lack of people wishing to use our developments due to the lack of sane use cases. Until 2014 the domestic market worked in the paradigm of G2M (“goods to person”) that is, the desire to completely robotic warehouse.

Foreign manufacturers actively demonstrated at the exhibitions of various manipulators, automatic placement and withdrawal of the goods (AS/RS), conveyors and other automation equipment distribution, which began to be actively used in the warehousing. However, a sharp jump in the dollar and Euro made their adjustments, and the market has returned to technology M2G — “man to product”, — having turned their attention to people as a primary resource productivity. By this time we already had audited more than 50 customers, we came to the conclusion that even in those warehouses where the embedded management system (WMS), called the use of human resources is rational, does not work. For example, in many warehouses there is no load distribution.

Even the design takes into account only the marginal burden on the employee in the form of bulk or weight, and the order of operations does not appear anywhere. Thus, it is absolutely unequal division of labor. Someone can get job to work with heavy loads that require recovery after multiple transactions and someone will work with “Mellstock”, with low labor costs for operations, but high — passing route (“runs”). And we are again reminded of its analogy with the gym. Only by this time already became available for new gadgets such as fitness bracelets.

We combined their experience with the capabilities of the bracelets and went back to our customers with a proposal to test the technology on their capacity. It was about a warehouse with an area of about 8 thousand square meters, which has 19 employees, and second — more than 10 thousand square meters and 25 employees. The first tests showed that the metering pulse of different people produces different results. The pulse may be different at rest, and besides, people used to the workload, and the fact that a week ago seemed to be a feat, tomorrow becomes part of everyday work.

To interpret the data we connected to the output of the neural network and were able to create for each employee a private “physical profile”, which allows to interpret the results and make decisions about alternating tasks to increase productivity, avoiding “overload” of employees. With their results, we turned to professional doctors and medical consultants, it turned out that with the right approach and monitoring of the health condition of the employee, alternating load, we can easily increase productivity by 8-10%. Today we use the bracelets, which can receive information on the number of steps taken and heart rate of the employee, but are waiting for new gadgets that can transmit data about the pressure, saturation of blood oxygen and other parameters, and this will help to improve the performance is already at 30-40%. To achieve such performance is possible, but you have to understand that no technology will replace the culture of work. Over the years, we observed the difference in approach to the organisation of warehouses of the Russian and foreign customers.

Someone chooses the path of cheaper and generic products, someone who relies on physically strong employees — “bullies”, someone sees a way in technology. In Europe and the United States has long been a warehouse not associated with low qualifications and poor wages, this is a work where staff hold years of passing of professional retraining, including the training of competent physical stress. For Russian business it is not tradition, and it is those technologies that will not only change the approach to the process, but also seriously reflected in the financial statements, can become the catalyst for change. For example, the salary of warehouse worker is 30 thousand rubles (as of today it is not the highest indicator) for the company, this amount, of course, above — round up to 40 Grand a month. Again, lets say that the warehouse is 20 people.

Technology needs in the district three months (this data gathering, analysis, configuration, and so on). If you even lay a achievable result in 10% growth performance, get an average of 4 thousand rubles a month cost savings on a single employee, and about 80 thousand a month for the whole team. For the year at rates of 10% is about one million rubles. The more employees or the higher their salary, the more the number of. During the year the technology pays for itself completely, even for small enterprises — be it a warehouse, manufacturing, agriculture or retail.

Leave a Reply