PlanningVis: A Visual Analytics Approach to Production Planning in Smart Factories

Dong Sun, Renfei Huang, Yuanzhe Chen, Yong Wang, Jia Zeng, Mingxuan Yuan, Ting-Chuen Pong, Huamin Qu

Abstract

Production planning in the manufacturing industry is crucial for fully utilizing factory resources and reducing costs. With the advent of Industry 4.0, the massive volume of data recording factory status presents an opportunity to understand and adjust production plans through a data-driven approach. However, developing a systematic analytics approach is challenging due to data volume, complex product dependencies, and unexpected market or plant changes. We propose PlanningVis, a visual analytics system that supports the exploration and comparison of production plans across three levels of detail: a plan overview, a product view, and a production detail view. By integrating an automatic planning algorithm with interactive visual explorations, PlanningVis facilitates efficient optimization and quick responses to unanticipated incidents. Two case studies and expert interviews demonstrate the system's effectiveness and usability.

Introduction

In modern smart factories, planners must navigate a "wicked" problem: balancing four competing objectives—minimizing order delays, reducing production costs, maintaining low inventory, and ensuring a "smooth" or consistent workload for machinery. Traditional approaches often fall into two inadequate camps: purely automated "black box" algorithms that lack real-world flexibility, and manual spreadsheet-based planning that is slow and error-prone. PlanningVis bridges this gap by combining the speed of mathematical optimization with human intuition.

The PlanningVis System

PlanningVis utilizes a hybrid backend and a three-tiered visualization frontend to manage the complexity of modern manufacturing.

  • Hybrid Algorithm: The system uses a combination of Linear Programming for optimization and heuristic algorithms to ensure computation remains fast enough for interactive use.
  • Multi-Level Visualization:
    • Plan Overview: Employs custom "plan glyphs" to visualize the overall health and performance of different plan versions.
  • Product View: Features a "Segmented Parallel Coordinates Plot" to handle high-dimensional product data and a gallery of product glyphs to identify outliers.
  • Production Detail View: Displays the Bill of Materials (BOM) as a tree structure linked to daily production heatmaps, allowing planners to trace delays back to specific component shortages.
The PlanningVis interface featuring (a) Configuration Panel, (b) Plan Overview, (c) Product View, and (d) Production Detail View.

Key Visual Innovations

Visualizing Variance

A core innovation in PlanningVis is the design of product glyphs that go beyond average performance indicators. These glyphs use the angle of an arc to represent the variance of daily indicators. This helps planners identify products that may appear stable on average but suffer from high daily volatility, which could indicate underlying production instability.

The product glyph design (a) and its alternatives (b, c), showing performance indicators and their variances.

Human-in-the-Loop "What-If" Analysis

The system is designed for interactive testing. Planners can adjust constraints—such as simulating a plant shutdown by adding a "holiday"—and immediately see the ripple effects across the supply chain. This juxtaposition of the BOM tree and production heatmaps reveals hidden constraints, such as when a finished product is delayed because shared raw materials were prioritized elsewhere.

Case Studies and Results

The system was evaluated using a real-world dataset from a global telecommunications manufacturer, covering 1,038 products over a 30-day period.

  • Optimization Success: Planners identified a bottleneck in specific capacity sets. By increasing capacity by 50% in the simulation, they significantly reduced order delays without a prohibitive increase in production costs.
  • Crisis Management: During a simulation of Typhoon Mangkhut, planners added a 2-day shutdown to an affected plant. PlanningVis allowed them to quickly re-allocate resources and mitigate the resulting delays.
  • Expert Validation: Domain experts noted that the system significantly improved efficiency over traditional table-based workflows, particularly in localizing the root causes of production issues.

Conclusion and Future Work

PlanningVis demonstrates the power of combining automated optimization with interactive visualization. While the current system provides significant improvements, future work will focus on reducing algorithm latency for even larger datasets and incorporating logistics costs and transportation times between factory locations to provide a more holistic planning tool.

Summary updated

Abstract

Production planning in the manufacturing industry is crucial for fully utilizing factory resources (e.g., machines, raw materials and workers) and reducing costs. With the advent of industry 4.0, plenty of data recording the status of factory resources have been collected and further involved in production planning, which brings an unprecedented opportunity to understand, evaluate and adjust complex production plans through a data-driven approach. However, developing a systematic analytics approach for production planning is challenging due to the large volume of production data, the complex dependency between products, and unexpected changes in the market and the plant. Previous studies only provide summarized results and fail to show details for comparative analysis of production plans. Besides, the rapid adjustment to the plan in the case of an unanticipated incident is also not supported. In this paper, we propose PlanningVis, a visual analytics system to support the exploration and comparison of production plans with three levels of details: a plan overview presenting the overall difference between plans, a product view visualizing various properties of individual products, and a production detail view displaying the product dependency and the daily production details in related factories. By integrating an automatic planning algorithm with interactive visual explorations, PlanningVis can facilitate the efficient optimization of daily production planning as well as support a quick response to unanticipated incidents in manufacturing. Two case studies with real-world data and carefully designed interviews with domain experts demonstrate the effectiveness and usability of PlanningVis.
BibTeX
@article{sun2020planningvis,
  author = {Dong Sun and Renfei Huang and Yuanzhe Chen and Yong Wang and Jia Zeng and Mingxuan Yuan and Ting-Chuen Pong and Huamin Qu},
  title = {{PlanningVis: A Visual Analytics Approach to Production Planning in Smart Factories}},
  journal = {IEEE Trans. Vis. Comput. Graph.},
  year = {2020},
  url = {https://openreview.net/forum?id=jkQPn6tIqN},
}

Loading related content...

Figure