AI-based workforce planning: automated process and proactive planning through data power.

Artificial intelligence (AI) is transforming intralogistics and fundamentally changing everyday warehouse work. AI is automating and speeding up time-consuming routine tasks. We are already using AI to optimize storage and retrieval processes using the so-called "warehouse healing" strategy. The focus here is on shortening travel times, improving the warehouse structure and reducing throughput times during order picking. Nevertheless, the question arises: are these measures sufficient to fully exploit the potential for increasing efficiency in the warehouse? No, we say. Sustainable and forward-looking personnel planning is just as essential. Our data science expert Timofej Woyzichovski is therefore developing an AI-supported solution for personnel planning in intralogistics.

Behind the supposedly simple term "personnel planning" lies a time-consuming and demanding process. Specialist staff often determine and plan the ideal staffing for each warehouse area manually without precise knowledge of the volume of work. The legal and operational regulations as well as the qualifications and preferences of the employees must also be taken into account. The aim is to ensure the best possible customer supply and at the same time reduce the workload on employees. However, especially in complex and large warehouses with many employees and variable requirements, this process is very error-prone. This is a task that AI can take over in the future and simplify personnel planning at the touch of a button.

From theory to practical solution
Woyzichovski supervised one of our employees on her bachelor's thesis. The topic was: Assessing the increase in efficiency through the implementation of smart devices, automation and data science in order picking. "By analyzing the existing data from our warehouse management software, such as past incoming orders and stock levels, as well as taking historical log data into account, I recognized the great potential for the further use of AI," says Woyzichovski. Despite careful planning, unpredictable order peaks, operational changes at short notice and staff absences present companies with major challenges. AI can recognize patterns and trends in the data and predict when peaks are to be expected. The aim is to use intelligent workforce planning to plan and act proactively rather than just reactively.

Advanced technology sets the pace
Woyzichovski is developing an AI-based solution that predicts staffing requirements in the warehouse and optimizes workflows. To do this, he uses data from the warehouse management software SuPCIS-L8 and the automatic storage system AutoStore of the customer Hermann Müller Elektrogroßhandel GmbH. Complex models from the field of time series forecasting are used to ensure the accuracy and reliability of the forecasts. Hermann Müller is very open to advanced technologies. The company hopes that the use of AI-based workforce planning will give it a competitive edge. They also want to manage the operational challenges in their warehouse operations more effectively and reduce operating costs.

What can AI-based workforce planning do? The capabilities at a glance.

1. The AI analyzes historical data such as order volumes, seasonal fluctuations and sales promotions to identify patterns and trends. It uses this information to forecast future workloads.
2. It precisely plans the number of employees per shift. This avoids overstaffing and understaffing, which causes overtime or unused resources.
3. The AI automatically creates the optimal shift plan. In doing so, it takes into account the predicted workload, the availability and qualifications of the employees as well as legal and company regulations such as working hours and rest periods.
4. AI-based personnel planning reduces the administrative effort involved in shift planning. This gives specialist staff more time for other important aspects of warehouse operations.
5. The AI creates balanced and transparent shift plans that take individual preferences and qualifications (if documented) into account. This promotes motivation, ensures job satisfaction and helps to increase productivity.
6. With the help of AI, specialist staff can make decisions based on data instead of intuition or outdated methods. This leads to more objective and well-founded decisions in HR management.
7. If personal data is also recorded in the log data, it analyzes the performance data and identifies areas for improvement. It shows which teams work more efficiently at certain times. This leads to better planning.

Challenges during implementation
Ensuring data quality and availability is a major challenge. Incomplete data impairs the performance of the AI and outdated data distorts the results. To prevent both, we have built in an additional data check and cleansing function. The integration of AI-based personnel planning into existing IT and personnel management systems is complex. It requires seamless interfaces for data exchange. Another challenge could be concerns on the part of the workforce. Especially if there are fears that AI-based personnel planning will replace human workers. It is therefore important to create acceptance through transparent communication and training. Modeling AI for workforce planning purposes is complex. There are many variables to consider and constant adjustments to changing conditions are necessary. Compliance with legal and ethical framework conditions, particularly with regard to data protection and the handling of employee data, must be guaranteed. Involvement and continuous exchange with employees are essential. The AI tool must also be scalable in order to adapt to changes and company growth, which this solution is.

Universal application options
The data analysis from the AutoStore at Hermann Müller can also be applied to other warehouses and customers. This applies regardless of which warehouse technology they use. The solution is primarily based on data from the SuPCIS-L8 warehouse management system. The processes and algorithms developed for data analysis and personnel planning can be flexibly transferred to different warehouse configurations and conditions as long as the same data types are available. This enables broad implementation across different warehouse locations and systems.

‍Practicaltests and future developments
AI-based personnel planning supports specialist personnel in intralogistics. It improves the quality of customer care and employee satisfaction. However, the expectations of AI-based personnel requirements forecasting should be realistic. AI provides suggestions, but does not replace a final assessment by experienced specialist staff. The solution will be tested under real conditions at Hermann Müller in fall 2024. This will allow us to optimize the solution based on the feedback and specific requirements. Further functions are planned in the second stage. These include the ability to incorporate short-term changes such as sick leave or order peaks into the calculation of personnel planning. AI-based personnel planning can be used to quickly reschedule and adjust shift plans. This keeps the warehouse optimally staffed, minimizes downtime and improves service quality.