Problem:
Mahindra & Mahindra previously used Excel macros and manual SAP data extraction for warehouse planning, which led to long processing times, data errors, complexity issues, and limited real-time collaboration.
Solution:
The digital transformation team used AnyLogic to build a simulation model for better space utilization and overall warehouse optimization. The model replicated the processes, including importing cleaned SAP data, calculating space consumption, applying logical constraints based on storage types, and simulating part handling at storage locations.
Results:
- Streamlined and provided reliable statistical analysis.
- Reduced planning time from days to 10–15 minutes.
- Improved accuracy in space forecasting and part categorization.
- Facilitated seamless comparison of storage strategies.
- Achieved a data-driven feedback loop.
Introduction: digital transformation of India’s leading vehicle manufacturer
Mahindra & Mahindra (M&M) is India’s leading vehicle manufacturers. It operates across various verticals including automobiles, farm equipment, electric vehicles, etc. This case study was developed in collaboration with Mahindra’s Auto Digital Centre (ADC) and Digital Supply Chain Management (DSCM) & Warehousing team.
ADC is a digital transformation team under M&M’s automobile division based in Pune. The team collaborates with R&D and production units to digitize and optimize core manufacturing processes using advanced technologies such as simulation, virtual reality, and digital twins.
The DSCM & Warehousing team focuses on planning, sourcing, delivering, and storing. Their specific mandate is to ensure adequate warehouse space utilization for all automobile part types, meet average daily demand, and forecast future storage requirements.
Problem: high lead time and complexity of processing storage setups
The DSCM & Warehousing team approached the ADC team to address critical warehouse optimization challenges. Previously, warehouse space planning relied heavily on Excel macros and manual SAP data extraction. This method posed several challenges:
- High lead time for processing and validating storage setups.
- Difficulty managing complexity and large volumes of data.
- Risks of data errors and limited collaboration with real-time systems.
- Lack of flexibility in evaluating alternative storage configurations.
As M&M scaled up production and warehouse complexity, it became essential to modernize the planning process to support faster, more accurate decision-making and long-term warehouse optimization.
Solution: a simulation model for warehouse optimization
The goal of this initiative was to conceptualize and evaluate the warehouse space utilization to handle demand-driven storage needs effectively.
The objective was to validate the space requirements derived from spreadsheet data, determine if current capacity met projected demand, and identify the most space-efficient storage type for each part. In addition, the model aimed to support statistical analysis of fluctuating part inflows and outflows to better understand warehouse behavior over time and enhance warehouse space utilization strategies.
The model prioritized two key factors: storage type and box type. Parts were categorized by constraints like height, box type, and stack limits. For example, corrugated boxes were stacked twice, metal pallets and PP boxes up to 2.7m, plastic bins up to 1m, and trolleys were non-stackable.
The ADC team turned to AnyLogic to build a simulation model to recreate the warehouse space utilization and achieve better warehouse optimization. The model replicated the following processes:
- Importing cleaned SAP data on parts, storage types, dimensions, stack limits, and forecasted demand.
- Calculating space consumption across three main storage types: Floor, High Bay Racks (HBR), and KLT bins.
- Applying logical constraints based on box types (e.g., corrugated boxes, metal pallets, PP boxes) and stackability.
- Simulating part arrival by truck, unloading, and internal transport via material handling equipment to storage locations.
The model allowed users to adjust key input parameters such as type of storage, pallet sizes, storage dimensions, and initial inventory—making it a flexible decision-support tool for different setups in ongoing warehouse optimization.
Why AnyLogic?
M&M had previously used other simulation solutions but faced limitations in addressing the full complexity of this warehouse optimization challenge.
Unlike static Excel tools, AnyLogic offered a more dynamic and robust approach. It allowed the team to test different warehouse space utilization scenarios, validate data in real time, and significantly reduce the need for manual data manipulation. The platform also supported fast, repeatable simulations for continuous decision-making.
Results: optimized warehouse planning and utilization
The simulation project resulted in several notable benefits:
- Reliable statistical analysis for warehouse space utilization. The analysis is independent of variations in part line items, facility locations, or resource configurations.
- Reduction in planning time from days to minutes. Once SAP data is formatted, simulations can be run and analyzed within 10–15 minutes.
- Improved accuracy in space forecasting and part categorization. The model accounted for diverse variables such as box dimensions, stack limits, and part flow volatility. This allowed for more precise planning and minimized errors in storage allocation.
- Integration of seamless comparison of storage strategies (e.g., Floor vs. HBR vs. KLT). The comparison helps to make better-informed decisions in warehouse optimization.
- Data-driven feedback loop of the warehouse simulation. The model generates optimized storage configurations that can be exported back to DSCM master data for validation.
The model also enabled the DSCM team to challenge prior assumptions and uncover inefficiencies in their legacy storage planning. For example, parts previously stored on the floor could be moved to HBR for improved storage efficiency and warehouse space utilization.
Following the successful first phase, the team plans to integrate the AnyLogic model directly with SAP. This will eliminate manual data handling and enable near real-time decision-making for warehouse optimization across operations.
The case study was presented by Arun Kumar G. from Mahindra & Mahindra at the AnyLogic Conference 2024.
The slides are available as a PDF.
