The idea behind the Voice technology is that the dialog between WMS and a user who perform warehouse operations (picker, stockman) is not a sequence of screen or print forms of WMS commands and information keyed or scanned (bar codes) by the user, but an exchange of verbal commands in the mother tongue of the user. Being integrated to a human-speech recognition system and a human-speech synthesizer, WMS is enabled to send commands to the user in a speech format and to receive back information entered as verbal commands. Information exchange between WMS and the user takes place on a speech level, as a result of which the hands of the user are freed from holding input devices and the eyes are freed from following a PC monitor hence the user can concentrate on performing an operation at hand.
The purpose of such large-scale testing in warehouse production was to establish the benefits of VT as compared to the traditional methods by using it in a limited production area, as well as to create technological conditions for implementing VT in all warehouses of the company.
Based on the results of preliminary studies, benchmarking world practices of implementing VT and prior negotiations, the company short-listed preferred partners for implementing VT.
To make the final decision on selecting the technology provider, the following requirements took precedence:
- Approved middleware technology to connect WMS to VT
- Experience implementing VT in Russia (completed and successfully functioning projects)
- Continuous technical support for middleware and VT applications.
As a result, only one provider met all the above requirements - Finnish Optiscan.
To have an unbiased comparison between the two technologies, the company decided on the following conditions of testing:
- Tests were held within a limited production area - the loft (piece goods storage); involving only one operation - picking piece goods.
- Two pickers were assigned to pick with Voice terminals and two other pickers - with traditional radio terminals. Half-way through the testing the pickers were supposed to switch the technologies to continue. This would provide objective results.
- The pickers were additionally motivated to achieve the optimal speed of work while adhering to the quality standards by being promised, as an incentive, a financial reward for the best results of work.
- The testing lasted for 5 weeks.
During this time the data have been collected on 7414 picks using VT, and 7464 picks using "classical" technology i.e. radio terminals. It has been established that the average time of picking with VT was 31 seconds, whereas using the "classical" technology - 34 seconds. The growth of productivity therefore was 9.7%.
The norm of 220 picks within 11 hour-shift as set in the beginning of testing proved to be realistic and often was exceeded. Thus the norms of productivity have been exceeded by nearly 70% to the record figures of 368 picks /shift for VT and 388 picks /shift for radio terminals!
Total quality control of performed tasks revealed that most of the differences were due to calculation errors. At that, the number of errors did not depend on the technology used and was practically the same for both methods (~0.5 % from the total number of picks).
To assess the efficiency of implementing VT on the basis of the results of the tests, the following has been calculated:
According to the statistics, 22 pickers on average worked with radio terminals in the tested production area in September 2011.
Therefore, the costs of implementing either technology on a full scale with the purchase of the necessary equipment would amount to:
The costs of implementing VT:
- Equipment (1 server + 24 terminals (2 - in reserve) + 48 batteries + 5 charging devices) = E1
- Software licenses (24 licenses) = 24 * l1 = L1
- Technology provider´s project costs = IE1
- Middleware technology project costs = IMW1
Thus the total cost is: CV = E1 + L1 + IE1 + IMW1
The costs of implementing "classical" radio terminals:
- Equipment (1 server + 24 terminals (2 - in reserve) + 24 additional batteries + 6 charging devices) = E2
- Software licenses (40 connections) L2
The total cost of implementing radio terminals is: CR = E2 + L2.
Thus, the difference in the costs of implementing VT and radio terminals is
ΔC = CV - CR
The savings of wage funds in a month can be calculated as follows:
The savings of wage funds = cost of work of one worker/hour х 24 hours х 30 days х 22 workers - cost of work of one worker/hour х 24 hours х 30 days х 20 workers (considering the growth of productivity by 9.7% that gives savings worth of 2 workers).
The savings of wage funds in the tested production area thus amounted to 273.600 roubles a month.
This means that the break-even point will be reached in 12 months after the deployment of the system.
Based on the results above, in the opinion of the project team VT is suitable for warehouses with big volume of piece goods picking (more than 60.000 pick tasks/month), also, VT should be tested on other operations, such as box picking.
While implementing the new technology at the warehouse the project team successfully tackled the following problems:
In the beginning of the implementation phase the warehouse staff had health safety concerns regarding wireless technology used by Voice terminals. However, this problem was resolved as soon as the technology provider made available all the necessary certificates of conformity.
The results of the first stage of the project showed that the workers were not motivated enough to switch to new technologies. However, when financial incentives were introduced, such as bonuses for best results of work, the productivity of VT pickers has increased 2,4 times. At that, the norms of productivity set earlier for work with the radio terminal have been exceeded by 76%. This proves that the productivity at the warehouse (as well as in the company overall) can be significantly increased by introducing financial incentives.
In the course of working with VT the workers agreed that they could pick more efficiently mentioning the following advantages of the new picking method:
- VT prompts the slot address step-by-step while the picker is en route to the picking location
- Hands-free equipment
- The item number confirmation step has been removed - this is the only change to the production process without any negative impact on the quality of work.
Resource: Skladskoy kompleks №1 January - March 2012