Features and innovations
Feb 2024
5 min.
Features and innovations
Rebecca got to the bottom of this question. She studied industrial engineering and wrote her bachelor's thesis on the topic with us:
Assessment of the increase in efficiency through the implementation of smart devices, automation and data science in our warehouse management software (WMS) SuPCIS-L8 in order picking.
This article looks at the results in detail, specifically how performance (picks per hour) can be increased by 23% through the use of smart devices in order picking.
Smart devices are: mobile computers, tablets, wearable computers or data glasses.
The devices are usually connected to the internet and are used for employee guidance in order picking. Relevant picking information is displayed to employees with the aim of maximizing picking performance and minimizing error rates.
A mobile data collection device (MDC device) from Zebra with an integrated scanner is used in the experiment.
In the practical section, conventional order picking without the use of technology is compared with technical order picking with the MDE device.
Two picking processes are described for this purpose. The one without the use of technology is based on a customer example. The one with an MDE device is based on practical experience, customer examples and the application functions of our warehouse management software.
The processes are developed graphically in the form of a flow chart. The processes form the basis for the qualitative and quantitative evaluation of the efficiency increase. The quantitative assessment criteria are derived from the results of the MTM method. The picking time and the productivity achieved are calculated.
The model test provides information about other influencing factors that cannot be determined by calculation. The quality of order picking is determined on the basis of the error rate and error type. This makes it possible to determine how experienced the employees are and how high the dependency on expert knowledge is. In addition, qualitative aspects such as the improvement of the work process, employee satisfaction and working conditions can be assessed.
The times determined in the test can be compared with the theoretical times of the MTM method. This comparison provides information on the human factor and gives an insight into the actual stability of productivity.
In the experiment, the shelf storage system corresponds to three shelves with 25 compartments each. One shelf is a storage area and all three form the picking area. The items are stored chaotically. It is possible that similar items are stored in the same compartment. In both cases, picking is carried out as a tour through the warehouse according to the multi-order picking principle.
Four customer orders are processed in one picking run. For single-stage picking, each of the customer orders is assigned a target container on the picking trolley. The pick list contains all items of the four customer orders.
The ideal number of customer orders or the number of picks per picking run was not calculated. The selected number is based on the given possibilities of the model test. The other assumptions regarding the order structure are also not based on any specific calculations.
The target times are determined using the MTM method. This involves reducing the processes to their movement elements, assigning times and combining them into a key figure using formulas. The qualitative aspects and the actual times are determined in the test.
Preparation: The warehouse management software must be set up for picking with the MDE device. In addition, the orders must be generated according to the order structure with the order number, article, article description and article quantity.
In conventional picking, the picking list consists of four papers. Each is a customer order. After the last removal, each paper is assigned to a target container. The sequence of the items to be picked is not based on route optimization.
Implementation: In both cases, an identical number of orders are picked. After each picking run, the times are recorded and the containers are checked for errors. The picking trolley is emptied after completion and the items are returned to their original storage location. This creates the same conditions for the next picking run.
Follow-up: As soon as all picking orders have been processed, the development of the times, the average times and the number and type of errors are analyzed. On this basis, the two processes are compared with the previously calculated target times.
In order to better place the described picking processes with and without smart devices in the overall context, the upstream and downstream activities are briefly described.
Conventional picking with pick list
The process is partly based on a customer example. The employee receives the individual customer orders (pick list), a pen and a picking trolley with four bins.
Upstream activities: Customer orders are received by e-mail, printed out, checked for urgency and color-coded accordingly. Orders that need to be processed promptly are highlighted in red. All orders are placed at a collection point. Picking trolleys with numbered containers are also available. The process starts at the collection point.
Picking: The employee takes four customer orders, starting with the ones marked in red, and a picking trolley from the collection point. He goes to the first storage area and checks the items on the pick list for a withdrawal. If there is a withdrawal, the pick list shows the storage location, the article and the quantity. He then removes the goods and places them in the target container assigned to the customer order on the picking trolley.
The removal is confirmed by ticking off the item. If there are more items to be picked in this storage area, the process is repeated. If there are no further withdrawals, the employee moves to the next storage area. All items in this storage area are checked again for removal. The process is repeated until all items from all storage areas have been processed.
The pick lists are placed in the corresponding containers for further identification and the picking trolley is parked at the collection point. The collection point is the destination or end of the picking process for the picking trolley.
Downstream activities: The filled order picking containers are taken to the packing station and prepared for dispatch by other employees.
Order picking with smart devices
The employee receives a picking trolley with four bins and the mobile data capture device (MDE device) from Zebra.
Upstream activities: Outside of setting up and preparing the picking trolley, only the login with user name and password on the MDE device is required. The employee selects their outgoing goods activity with the respective stored category of stock removal.
Customer orders are compiled into picking orders and prioritized by the WMS. This optimizes the routes and intelligently adjusts the sequence of the items to be picked. Sorting by storage location is based on the numerically ascending labeling of the storage bins in the system. (Based on this logic, other picking run patterns can also be implemented in the WMS).
Picking: The employee takes a picking trolley and scans the barcode of the target container numbers. The MDE device displays the picking order, which starts with a click. The dialog with the picking information on the storage location, article and target container opens. The relevant information is displayed visually and the data field to be confirmed is highlighted in color. Only the next item to be picked is displayed. This reduces the employee's cognitive workload and allows them to concentrate on the picking task at hand.
The employee takes the storage bin from the display and scans the barcode there. The WMS checks whether the barcode information matches the expected entry. If this is the case, the item information is displayed. If an incorrect storage bin has been scanned, an error message appears and the process must be repeated. The same principle applies when picking the required item.
The item information is highlighted in color and the employee is prompted to scan the item. The system checks whether the information is correct and displays an error message if necessary. If everything is correct, the quantity is entered. This is simply confirmed with a click. The target container is then marked, which must be confirmed by a scan. The WMS checks whether the data matches the information in the order. The removal is now complete.
If the WMS detects that there is another removal, the corresponding dialog is automatically called up and displayed on the MDE device. The process starts again from the beginning. If there is no further removal, picking is complete. The employee receives a notification of completion and the picking trolley can be parked at the collection point.
Downstream activities: As in the conventional picking process, the items combined according to customer orders are taken to the packing station for preparation for dispatch.
The evaluation and comparison using the MTM method shows that picking with smart devices is 17% faster than conventional picking with a pick list.
Conventional picking: picking time of 38.7 seconds per pick, 62 items per hour or 93 picks per hour
Picking with smart devices: Picking time of 32 seconds per pick, 75 items per hour or 112 picks per hour.
This means an increase in efficiency of 21% (92.85 vs. 112.1 picks per hour).
The differences lie in the processing time, which includes the base time and the dead time. The differences are particularly noticeable in the base time with a difference of 43% (6.19 vs. 3.52 seconds). The dead time also has an influence with a delta of 19 % (38.75 vs. 31.45 seconds).
The 2.67 seconds of base time saved by the technical process corresponds to almost half the base time of the conventional process.
In the conventional process, the four picking orders must be identified and picked at a collection point (approximately seven meters away from the picking trolley collection point). In addition, at the end of the picking run, the corresponding page of the picking list must be attached to each picking container.
In the technical process, only the MDE device needs to be taken, the picking trolley scanned and the process started with a click. The reduced effort at the start of the technical process is noticeable in the base time.
The delta of 7 seconds in the dead time can be attributed to three reasons.
At first glance, the efficiency of the test results appears sobering compared to the results with the MTM method. With an average picking time of 32.6 seconds per pick, conventional picking is 9% faster than the picking time of 35.6 seconds per pick with smart devices.
Conventional picking is therefore 6 seconds faster per pick than calculated using the MTM method. Picking with smart devices, on the other hand, is 3.5 seconds slower than calculated. The discrepancy of 3.5 seconds between the test results and the calculations of the MTM method can be explained.
In the test, the same picking trolley was used for each picking run. However, this cannot be mapped in the WMS. The picking trolley is considered occupied until the time at which the picked items are completely packed. To avoid this problem, a new picking trolley would have to be stored in the system for each picking run. However, this means that the target containers receive a new number with each picking run and each new picking trolley. The attached barcodes are then no longer universally valid and the 7-digit container number would have to be entered manually for each removal. 3.5 seconds is an expected value for this process. Without the additional effort described above, a time of 32 seconds per pick would be expected. This would make the picking time identical to the conventional picking time determined in the test.
The difference in conventional picking of 6 seconds per pick in the test compared to the MTM method can be explained by the human factor.
As all preparations and storage activities were carried out by the employee shortly before the test, the employee is considered experienced. Reading the pick list, searching for the storage location, the articles and the target container is significantly faster than assumed in the MTM calculation.
The time in the first two conventional picking runs is significantly higher than the average picking time. From the third picking run and the experience gained, the times improve.
At approx. 41 seconds per pick, the second picking run is close to the calculated 39 seconds per pick. The results of the MTM method correspond to the performance of an employee with less experience. With the exception of the fifth picking run, in which there was an inventory error, the times up to the seventh picking run are significantly better than the average. In addition, the employee improves by sometimes picking two items at the same time, which was not taken into account in the MTM method. From the eighth picking run onwards, concentration wanes and performance stagnates slightly above average.
The blue curve shows that performance in conventional picking is heavily dependent on people. It was perceived as laborious by the employee and concentration quickly wanes.
With the orange curve, neither a familiarization effect nor a learning effect can be observed. The employee starts with a very good performance level from the very first order picking.
In contrast to the stagnating performance depending on the concentration level in the conventional process, an irregular progression can be observed in technical order picking.
The above-average performance in picking runs three, ten and thirteen can be attributed to the software-based route optimization.
The peak in the eighth picking run was caused by a distraction. In the twelfth picking run, the employee reported a loss of concentration. However, the loss of concentration occurs much later with smart devices. The employee confirms that picking with smart devices is less strenuous and requires less concentration. The repetition of a scan or removal that may be necessary due to a correction note from the WMS can also be a reason for a longer picking time.
The seamless employee guidance and the system-side control processes lead to a significantly better error balance than with conventional picking. During the 15 picking runs, there were seven errors in conventional picking and only one error in picking with smart devices. On average, one error occurred in every second picking run in the conventional process.
The most common errors were mixing up quantities with the number of the target container and forgetting to remove an item due to a lack of overview. An error in the restocking process meant that the item could not be found in the specified storage bin.
Smart devices can be used to react quickly and correct stock levels. The recurring scanning process and comprehensive networking enable seamless inventory management for maximum transparency and flexibility in the warehouse.
The only error was the use of an incorrect target container. However, the error only occurred because the target container was entered manually. If the target container had also been confirmed by scanning, the error would have been detected and prevented at an early stage.
Errors in order picking are serious. If they are not noticed in downstream processes, they reach the customer. This leads to dissatisfaction, returns and additional costs. Even if errors are noticed in subsequent processes, this leads to delays and additional work, which again increases costs. Minimizing errors is an important part of increasing efficiency.
Overall, picking with smart devices achieves better results:
The model test is representative of order picking by an experienced employee. The results calculated using the MTM method apply to employees without experience.
For inexperienced employees, the introduction of smart devices leads to a 17% improvement in picking time and a 21% increase in picking performance. At the same time, the error rate is significantly reduced. For experienced and very good employees, performance remains at the same level with fewer errors and more transparency.
Overall, the calculations and test results confirm an increase in efficiency through the use of smart devices.
Although the results obtained in the test environment are easily comparable, they do not correspond to reality. The walking distances in the model test are too short. To eliminate these differences, the routes were simulated based on the time proportions determined by Pulverich & Schietinger (2009). According to this, the walking time is 55% of the total picking time.
In the MTM calculation, the merging time represents only 15% or 18% of the total picking time due to the very small shelf system under test conditions. If the merging time in the MTM calculation is increased to 55% of the picking time, the picking time per pick is 54 seconds in the conventional process and 44 seconds with smart devices. Calculated backwards, this corresponds to a picking path of 40 to 50 steps per order item.
The simulation results in a new picking performance of 66 picks per hour using the MTM method with the conventional process and 81 picks per hour when using smart devices.
This corresponds to a new efficiency increase in picking performance of 23%. These values form the basis for assessing the efficiency increase through automation in the next step.
The introduction of warehouse management software such as our SuPCIS-L8 is required for the use of smart devices as the first digitization step.
The Warehouse Management System prepares the information, filters it and makes it available to the employee on the MDE device. Only the data relevant to the current activity is displayed simply and legibly. The WMS also has a monitoring function.
All control and decision points are made by the WMS. Interaction with the WMS takes place at every data point. Through communication and networking, the WMS generates trust and security on the one hand and transparency on the other, which increases efficiency.
Changes to inventory data due to stock transfers and removals or integrated inventory corrections are transmitted in real time. Inventory checks are carried out daily so that errors can be detected and corrected in good time. This ensures seamless inventory management and inventory security. End-to-end traceability is guaranteed.
Real-time stock levels and the continuous availability of data create a certain degree of flexibility. For example, they enable urgent orders to be processed at short notice and a rapid response to unforeseen events.
The WMS increases process reliability and enables optimum utilization of all warehouse resources. This reduces storage costs and increases efficiency.
Automatic identification systems, such as optical barcodes as used in the test, can be implemented across the board. The software makes this compressed information readable for both machines and humans.
Once this stage of digitalization has been implemented in the warehouse, the next development can be tackled with the aim of increasing efficiency: Automation.
Notes: The origin of the data and information used to investigate the use cases is partly theoretical and partly practical. The studies also refer to different use cases with different methodologies. As a result, the results are not exactly comparable with each other and are not universally valid.
In addition, efficiency in order picking depends on many factors. For example, the process characteristics, the warehouse topology, the order structure and the interaction of technologies. Efficiency is correspondingly variable depending on the application.
Nevertheless, the approach chosen for this work is still the best one to answer the leading question. This is because the respective efficiencies of picking can be analyzed and related to each other, taking into account the differences. Thus, the findings obtained are complete, valid and plausible when the underlying overall concept is taken into account.
Blog post with an overview of the overall results: Amazing results in increasing efficiency through smart devices, automation and data science.
Blog post with the automation results in detail: Automation increases efficiency by 97% (semi-automated) and 140% (fully automated) in order picking.
Blog post with the data science results in detail: Reduce outsourcing routes by 20-25% and optimize efficiency through intelligent algorithms.
Blog post: Outlook for future developments: Increasing efficiency in intralogistics.
