#28: Nesting in 3D Printing, Multilingual Customer Service, and Withholding Containers
Hello AI Enthusiasts! π
Welcome to Issue #28 of Good AI Vibes.
In this issue, we delve into the powerful role of AI in shaping the future:π
π The Future of Fabrication: AI-Driven 3D Printing
π Around the Clock with AI: Optimizing Global Customer Success
π’ Efficiency Uncontained: AI's Role in Reducing Shipping Costs
Know someone passionate about leveraging AI in their sector?
Letβs embark on this journey together.
The Future of Fabrication: AI-Driven 3D Printing
What's the critical hurdle?
Imagine Liz, the Production Director at a leading manufacturer of plastic components that utilizes 3D printing technology. Liz faces a significant challenge: efficiently organizing the production schedule to handle a variety of uniquely shaped parts. Each part has specific timelines from when it needs to start being produced to when it must be finished. The stakes are high as any delay in producing parts can lead to missed deadlines, dissatisfied clients, and financial penalties. The complexity of arranging these irregular parts optimally on the print beds to maximize the use of space and minimize production time makes this task daunting.
How was this initially tackled?
Previously, Liz's team used standard software solutions designed for production planning. These programs could arrange parts in batches and assign them to available 3D printers. This method focused on placing as many parts as possible within each print cycle, using basic shapes to approximate the space each part would occupy. An example of this approach might be using rectangular bounding boxes to represent each part, regardless of its actual shape, to simplify the arrangement process.
Why did the initial approach fall short?
The traditional methods weren't sufficient because they couldn't adequately handle the complex shapes and specific timing needs of each part. The software often wasted valuable printing space due to its rudimentary shape approximations, leading to fewer parts per batch and increased production times. Moreover, this approach lacked the sophistication needed to prioritize parts based on their due dates, resulting in frequent tardiness in order deliveries. As a result, despite having a seemingly efficient system, Liz found that the production often ran behind schedule, impacting the company's ability to meet client demands effectively.
How did AI revolutionize the solution?
The introduction of AI revolutionized Liz's production planning by implementing a system that could intelligently consider each part's unique shape and timing requirements. The AI solution inputs the dimensions, release dates, and due dates of parts, and uses advanced algorithms to optimize how these parts are nested together on the print beds and scheduled for production. This method ensures maximum space utilization and timely production without manual intervention. For instance, the AI might determine the best way to interlock a zigzag-shaped part with a circular part to minimize wasted space. This optimization has led to a significant reduction in production times and an increase in on-time deliveries. Liz observed that with the AI system, the company could increase its throughput by 20% and reduce tardiness by 40%, enhancing customer satisfaction and operational efficiency.
Main reference: Nascimento, P. J., Silva, C., Antunes, C. H., & Moniz, S. (2024). Optimal decomposition approach for solving large nesting and scheduling problems of additive manufacturing systems. European Journal of Operational Research.
See other references at the end.
Around the Clock with AI: Optimizing Global Customer Success
What's the critical hurdle?
Meet Eric, the Chief Operating Officer at a global cybersecurity SaaS company that caters to clients across various time zones and languages. His main challenge is creating an efficient and effective schedule for the customer success team that ensures 24/7 coverage while accommodating language diversity. The complexity arises from the need to predict fluctuating customer demands across different regions, which can vary significantly due to factors like regional cyber threats or local business hours. For instance, if there is an unpredicted rise in cybersecurity threats in Spanish-speaking regions, the team must be prepared with enough Spanish-speaking agents ready to respond. Inadequate coverage could lead to slow response times, potentially leaving client systems vulnerable and damaging the companyβs reputation.
How was this initially tackled?
Originally, Ericβs team used a standard scheduling approach, assigning shifts based on historical trends of customer interactions, segmented by time zone and primary languages spoken. This system aimed to pre-determine the most likely times for increased demand and match the workforce accordingly. For example, if data indicated high request volumes from Asia in the early morning hours, the corresponding night shift in the U.S. would include agents proficient in Asian languages.
Why did the initial approach fall short?
This traditional scheduling model lacked the agility to adapt to sudden changes in demand, as it relied heavily on historical data that did not always reflect real-time market dynamics or emerging threats. As a result, the company sometimes found itself either overstaffed during off-peak hours or, more critically, understaffed during unexpected spikes in demand. This rigidity in scheduling often led to suboptimal use of resources and hindered the companyβs ability to provide timely customer support, especially in a crisis.
How did AI revolutionize the solution?
Eric introduced an AI-driven demand forecasting tool tailored to enhance scheduling efficiency while respecting fixed shift patterns to comply with employee rights. This AI tool analyzes various data points, including real-time customer interaction trends, historical data, and predictive analytics on potential cybersecurity threats by region and language. The system then forecasts the volume and nature of customer requests expected for each time zone and language group, enabling Eric's team to create a more informed and responsive scheduling plan for the upcoming weeks. For example, if the AI predicts a rise in requests from German-speaking regions due to a potential security risk alert, the schedule for the following weeks is adjusted in advance to include more German-speaking agents during the critical periods. This proactive approach allows the company to maintain high responsiveness and customer satisfaction, improving their operational efficiency and adaptability in a dynamic global market.
Main reference: Good AI Vibes research.
Efficiency Uncontained: AI's Role in Reducing Shipping Costs
What's the critical hurdle?
Imagine Richard, the Head of Supply Chain at a container shipping company based in Hong Kong. The company grapples with the costly and complex challenge of managing the storage and reuse of empty containers. Every day, containers are left idling at the hinterlands, incurring hefty transport and detention costs. Richard's dilemma is not just logistical but also strategicβhow to determine the right moment and quantity of containers to withhold for reuse, minimizing costs while maintaining service efficiency. This issue is paramount as poor container management directly translates into increased operational costs and reduced profitability, especially in a highly competitive shipping industry.
How was this initially tackled?
Traditionally, Richard's company tackled this by setting a fixed withholding threshold for containers based on historical data and general industry practices. This approach involved manually calculating the expected number of containers needed for future shipments and maintaining a constant buffer of containers at their facilities. The aim was straightforwardβkeep enough containers on hand to meet shipping demands without overstocking, thereby managing space and costs effectively.
Why did the initial approach fall short?
However, the conventional method fell short due to its static nature. It failed to adapt to the dynamic and often unpredictable changes in shipping demand and the availability of trucks and containers. This rigidity meant Richard often either ended up with too many idle containers during downtimes, driving up detention costs, or too few during peak times, missing out on potential revenue from shipping contracts. Moreover, this approach didnβt factor in the variations in individual consignee and shipper needs, making it inefficient in dealing with the complexities of modern container logistics.
How did AI revolutionize the solution?
The game changed when Richard's company integrated an AI-driven approach to optimize container withholding decisions. This AI system utilized data analytics to model the entire logistics as a double-ended queue where real-time data on container status, truck availability, and shipping demands were fed into the system continuously. By processing this data, the AI could predict more accurately when and how many containers should be withheld for optimal cost savings. It adjusted the withholding thresholds dynamically, considering current stock levels at the shipper, expected inbound and outbound container flows, and even nuanced exporter characteristics. The AIβs ability to adapt in real-time to the operational landscape not only reduced transport and detention costs significantly but also enhanced overall logistical efficiency. For instance, in a scenario where sudden demand spikes, the AI could instantly recalibrate its parameters to ensure enough containers were available, thereby averting potential revenue loss. The impact was profoundβa more resilient, responsive, and cost-efficient supply chain, primed for the complexities of modern trade.
Main reference: Legros, B., Fransoo, J., & Jouini, O. (2024). How to optimize container withholding decisions for reuse in the hinterland? European Journal of Operational Research.
See other references at the end.
Good AI Vibes Cartoons π
![](https://substackcdn.com/image/fetch/w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F22d111a8-aa48-485b-b58b-baf159ebda67_2008x852.png)
In case you missed our last edition, catch up on all the insights from Good AI Vibes #27 right here!