Helping inventory managers to prevent possible stock-outs using digital twin AI
Visualising future of FourKites in coming years using AI in Inventory management
My Role
Research, Benchmarking, Brainstorming, Interaction, Prototyping
Product
FIN AI at FourKites
Duration
6 weeks
Team
Worked with a Senior UX designer Rohit Menon
Impact
Reduced stock-outs, cost savings, improved order fulfillment, and time savings.
Impact
Improved daily operations for dispatchers and attract new users seeking a more efficient, user-friendly logistics management solution.
My Role
Research, Brainstorming,Interaction, Prototyping, Usability testing
Product
Yard Tool at FourKites
Duration
16 weeks
Team
Me(UX designer), Rohit Menon(Sr. UX designer),Austin Joerger(Product Manager), Dhawal Sharma(Product Manager)
In September 2023, during an AI workshop, we researched areas in the supply chain where AI could improve the user experience for future FourKites products. We discovered that users face issues in securing the right resources to prevent stock-outs, which can lead to significant losses. This realization sparked the initiation of this project.
Problem
Inventory managers face significant challenges in planning for potential stockouts. They struggle with carrier selection, transportation forecasting, supplier choices, and route optimization, which hinders effective risk mitigation and maintaining optimal inventory levels.
Solution
Our AI-powered solution employs digital twin technology to simulate potential order journeys. It provides detailed recommendations for suppliers, carriers, transport modes, and appointment times. By analyzing historical and real-time data, it offers proactive strategies to prevent stock-outs. Users can simulate different plans and choose the best one based on the impacts
Impact
This AI-driven solution reduces stockout risks by leveraging predictive insights and digital twin technology, enabling cost-effective and efficient decision-making for inventory managers. This proactive approach enhances supply chain efficiency and ensures optimal inventory levels.
Research
Contextual Inquiry
Understanding the current scenario of how stockouts are dealt with in inventories and identifying the present challenges and resources available to the users.
Stakeholder Research
After understanding how inventory managers deal with this issue, it was found that they have to use several third-party tools separately to make a single decision regarding potential outcomes. This involves a lot of analysis and decision-making.
Market Research
Inventory managers face challenges with existing tools, which often require separate usage for arriving at decisions. Here are some of the tools currently available in the market:
Kinaxis excels in supply chain planning with historical data but lacks advanced real-time data integration and simulation.
Logility excels in supply chain optimization and demand forecasting but lacks comprehensive risk analysis and advanced AI recommendations.
SAP excels in integrating supply chain processes and strong analytics but is hindered by high implementation costs, complexity, and limited user-friendly features.
Key Insights
Manual Stock Checks
Inventory managers spend significant time manually checking stock levels and predicting stockouts
Limited Visibility Tools
Lack of integrated tools providing end-to-end visibility and actionable insights.
Data-Driven Solutions Needed
Managers seek solutions integrating real-time data, historical trends, and predictive analytics.
Solution
Design Process of the Fin Sim(Simulation Dashboard)
Based on the user journey, mapped out possible scenarios where AI could be beneficial in addressing potential stock-out situations in the inventory.
From this, proposed several features.
_gif.gif)
FIN Sim (Simulation and recommendation tool)
We designed a comprehensive solution where FIN Sim(FourKites AI) sends alerts to users about potential risks. Users can then simulate solutions for better decision-making.
1
FIN Sim Alert system
FIN Sim(AI) alerts users about different exceptions in facilities globally.
2
Risk Selection and Simulation
Users select a risk to simulate the best solution as recommended by FIN Sim.
3
Recommendation Engine
AI recommends the best replenishment plans using historical data and future predictions.
4
Metrics for Decision Making
Each plan shows metrics indicating its profitability, helping users overcome decision paralysis.
5
Visual simulation of the order plan
AI visually simulates the order plan, showing origin, destination, route, supplier, carrier, and journey risks like future climate forecasts and other exceptions.
Reflection
Power of Predictive Analytics
Leveraging historical and real-time data revealed the power of predictive tools in preventing stockouts and optimizing supply chains.
Balancing Complexity with Usability
Designing an advanced yet user-friendly system taught me the importance of simplifying complex processes for intuitive use.
Adapting to Emerging Technologies
Using digital twin technology taught me how to integrate emerging tech into processes to enhance efficiency and decision-making.
Effective Communication
Presenting complex AI concepts clearly to stakeholders improved my ability to communicate technical details and benefits effectively.