arcelormittal delivers ai-enhances railway transportation ...arcelormittal delivers ai-enhanced...

3
ArcelorMittal Delivers AI-Enhanced Railway Transportation to Improve Efficiency and Reduce Costs Executive summary The demand for greater efficiency is a key driver of Industry 4.0—where analytics, artificial intelligence, and the Internet of Things (IoT) are transforming decision- making and productivity. A study of senior factory executives found that 86 percent reported major increases in shop-floor data collection over the past two years, and two-thirds reported that data insights have led to quality and efficiency savings of 10 percent or more. 1 A study conducted by Deloitte found that nearly all survey respondents—94 percent—reported that digital transformation is a top strategic objective for their organization. 2 ArcelorMittal, a major European steel producer, is improving efficiency and streamlining operations with machine learning technologies that are designed to accurately identify key data about railway cars that transport raw materials and ready products across factories and to the end customer. These valuable advancements are liberating employees from time-consuming and costly tasks. Challenges ArcelorMittal Poland moves large volumes of materials—with even smaller cargo weighing 20 or more tons. These materials, which are transported by railway cars, must be tracked along their journey. During a single day, there are over a thousand cars in motion at one factory. For each car, a variety of data points must be tracked, such as material location, the condition of the materials, and how fast shipments are moving. The company needed an automated solution that would liberate employees from needing to watch video feeds and input critical pieces of data. Not only was a scalable solution required, but also a solution that was durable enough to withstand harsh environmental conditions such as heat, rain, and cold temperatures. Solution ArcelorMittal Poland deployed a solution that leverages computer vision, deep learning, and near-real-time processing. The solution focuses on the identification and recognition of railway cars throughout the factory. The camera operators previously used video technology to categorize and tag incoming cars, which was a time-consuming and tedious process. For example, there are several checkpoints located throughout the site where railway cars are weighed. Operators must be available 24/7 to tag the cars, which adds up to a large number of work hours. Using the new technology frees up operators to focus on other important tasks. The new solution, which includes two cameras per checkpoint, has the capacity to log critical pieces of data quickly and accurately. For example, the railway cars pass through the weighing system and the algorithm, which uses the Intel Distribution of OpenVINO toolkit , matching the cargo weight to the image of the car. At a Glance ArcelorMittal and the Intel® Distribution of OpenVINO™ toolkit enable machine learning to identify critical data about railway cars and liberate employees from time-consuming and costly tasks. Deploy computer vision routines across the entire business. Deliver a multiplatform solution that scales up directly in IT-managed data centers. Accelerate performance and accuracy with the Intel Distribution of OpenVINO toolkit, which streamlines development of deep learning applications. “Intel supported us with knowledge, patiently answering our many questions, even on-site visits to help us with our code and make it smarter, faster, and ready to scale up, taking 100 percent of our processing capabilities.” —Gregorio Ferreira, data scientist and Industry 4.0 specialist Solution Brief Intel® AI: In Production AI for Industrial and Manufacturing Critical railway car data recognized and logged in near-real time with Intel® Distribution of OpenVINO™ toolkit and Intel® architecture

Upload: others

Post on 21-Mar-2021

3 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: ArcelorMittal Delivers AI-Enhances Railway Transportation ...ArcelorMittal Delivers AI-Enhanced Railway Transportation to Improve Efficiency and Reduce Costs Executive summary The

ArcelorMittal Delivers AI-Enhanced Railway Transportation to Improve Efficiency and Reduce Costs

Executive summaryThe demand for greater efficiency is a key driver of Industry 4.0—where analytics, artificial intelligence, and the Internet of Things (IoT) are transforming decision-making and productivity. A study of senior factory executives found that 86 percent reported major increases in shop-floor data collection over the past two years, and two-thirds reported that data insights have led to quality and efficiency savings of 10 percent or more.1 A study conducted by Deloitte found that nearly all survey respondents—94 percent—reported that digital transformation is a top strategic objective for their organization.2

ArcelorMittal, a major European steel producer, is improving efficiency and streamlining operations with machine learning technologies that are designed to accurately identify key data about railway cars that transport raw materials and ready products across factories and to the end customer. These valuable advancements are liberating employees from time-consuming and costly tasks.

Challenges ArcelorMittal Poland moves large volumes of materials—with even smaller cargo weighing 20 or more tons. These materials, which are transported by railway cars, must be tracked along their journey. During a single day, there are over a thousand cars in motion at one factory. For each car, a variety of data points must be tracked, such as material location, the condition of the materials, and how fast shipments are moving.

The company needed an automated solution that would liberate employees from needing to watch video feeds and input critical pieces of data. Not only was a scalable solution required, but also a solution that was durable enough to withstand harsh environmental conditions such as heat, rain, and cold temperatures.

SolutionArcelorMittal Poland deployed a solution that leverages computer vision, deep learning, and near-real-time processing. The solution focuses on the identification and recognition of railway cars throughout the factory.

The camera operators previously used video technology to categorize and tag incoming cars, which was a time-consuming and tedious process. For example, there are several checkpoints located throughout the site where railway cars are weighed. Operators must be available 24/7 to tag the cars, which adds up to a large number of work hours. Using the new technology frees up operators to focus on other important tasks.

The new solution, which includes two cameras per checkpoint, has the capacity to log critical pieces of data quickly and accurately. For example, the railway cars pass through the weighing system and the algorithm, which uses the Intel Distribution of OpenVINO toolkit, matching the cargo weight to the image of the car.

At a Glance ArcelorMittal and the Intel® Distribution of OpenVINO™ toolkit enable machine learning to identify critical data about railway cars and liberate employees from time-consuming and costly tasks.

• Deploy computer vision routines across the entire business.

• Deliver a multiplatform solution that scales up directly in IT-managed data centers.

• Accelerate performance and accuracy with the Intel Distribution of OpenVINO toolkit, which streamlines development of deep learning applications.

“Intel supported us with knowledge, patiently answering our many questions, even on-site visits to help us with our code and make it smarter, faster, and ready to scale up, taking 100 percent of our processing capabilities.”

—Gregorio Ferreira, data scientist and Industry 4.0 specialist

Solution Brief Intel® AI: In Production AI for Industrial and Manufacturing

Critical railway car data recognized and logged in near-real time with Intel® Distribution of OpenVINO™ toolkit and Intel® architecture

Page 2: ArcelorMittal Delivers AI-Enhances Railway Transportation ...ArcelorMittal Delivers AI-Enhanced Railway Transportation to Improve Efficiency and Reduce Costs Executive summary The

Solution benefits include:

• Simplified solution that enables computer vision routines across the business

• Ability to deliver a multiplatform solution and scale up directly in IT-managed data centers

• Greater performance and accuracy achieved through the Intel Distribution of OpenVINO toolkit, which allows developers and data scientists to accelerate and streamline the development of AI and deep learning applications and algorithms

ArcelorMittal Poland used the Intel Distribution of OpenVINO toolkit to run inference and deep learning models accelerated by Intel® FPGAs to realize greater efficiency.

Using GPUs was a consideration, but the company already had Intel® processors running its on-premise servers and could cost-effectively add in Intel® vision accelerators with Intel® Arria® FPGAs and the Intel Distribution of OpenVINO toolkit to accelerate inference. This allowed the company to keep the solution in IT-managed data centers, avoiding costly new infrastructure for development.

The Intel Distribution of OpenVINO toolkit has improved performance and allows the company to process up to 19 frames per second, compared to only two to three frames per second without the optimizations of the toolkit. What’s more, the solution takes far less memory and is running with less than 6 GB of memory, compared with 60 to 70 GB previously.

Solution Brief | ArcelorMittal Delivers AI-Enhanced Railway Transportation to Improve Efficiency and Reduce Costs

Make your vision a reality on Intel® platformsDevelop applications and solutions that emulate human vision with the Intel® Distribution of OpenVINO™ toolkit. The toolkit extends workloads across Intel® hardware to maximize performance:

• Enables deep learning inference at the edge.

• Supports execution across a variety of computer vision accelerators, including CPU, GPU, VPU, Intel® Neural Compute Stick 2, and FPGA, using a common application programming interface.

• Speeds up time to market via a library of functions and preoptimized kernels.software.intel.com/en-us/openvino-toolkit

Accelerate with developer toolsMore easily debug, analyze, build, and optimize on Intel platforms

Intel® System Studio provides a unified tool suite that simplifies the building of IoT solutions and embedded apps. software.intel.com/en-us/system-studio

Improve how you develop, test, and run your workloads

Intel® DevCloud for the Edge provides a cloud-hosted hardware and software platform for testing and optimizing on a cluster of Intel® hardware and software.software.intel.com/en-us/devcloud/edge

Prototype faster and expedite your path to productization

IoT developer kits and accelerators offer production-ready hardware preloaded with software.software.intel.com/iot

Explore and evaluate software

Download a wide range of free software tools from the Intel® Developer Zone to help you:

1. Get more from your code.

2. Maximize hardware capabilities.

3. Add competitive features by unlocking the unique technologies in Intel platforms.software.intel.com

Collaborate with others

Intel® AI: In Production is an ecosystem focused on reducing deployment complexities, promoting partner AI offerings, and increasing collaboration between its partners.intel.com/ai-in-production

Figure 1: Solution schematic

VMS proxy

Factory in remote location

Data centercluster

Cameras

Video managementsystem

Factory and logistic services

Intermediaryservers

Fast-responselocal services

Streaming API

Live video

Control room

DMZ

Triggers

Report

Video playback data Processing server

Intel® Distribution of OpenVINO™

toolkit

Web application

Website back end

Correction and acceptance UI

SQL 2019Machine learning services

Integration serverLocal applications

Production network domain

Cameras

2

Page 3: ArcelorMittal Delivers AI-Enhances Railway Transportation ...ArcelorMittal Delivers AI-Enhanced Railway Transportation to Improve Efficiency and Reduce Costs Executive summary The

Solution Brief | ArcelorMittal Delivers AI-Enhanced Railway Transportation to Improve Efficiency and Reduce Costs

About ArcelorMittal PolandArcelorMittal Poland is the largest steel producer in Poland and part of the ArcelorMittal group, the largest steel producer in the world. It concentrates about 70 percent of the Polish steel industry’s production capacity. The company is also one of the largest Polish exporters and producers of coke in Europe and in the entire ArcelorMittal group.

Learn more at poland.arcelormittal.com.

Conclusion ArcelorMittal Poland is pleased with the early results of the project as the company captures valuable insights and improves accuracy. By automating important processes, ArcelorMittal Poland can improve its ROI and is positioned to compete in the next generation of industry.

Learn moreFor more information about ArcelorMittal Poland, visit poland.arcelormittal.com.

For more information about the Intel Distribution of OpenVINO toolkit, visit software.intel.com/openvino-toolkit.

Figure 2: The Intel® Distribution of OpenVINO™ toolkit is a free software kit that helps developers and data scientists speed up computer vision workloads and streamline deep learning deployments from the network edge to the cloud.

“The Intel® Distribution of OpenVINO™ toolkit has evolved since the start of our journey. Now there are far more demos available that help you scale and build, and the documentation keeps getting better.”

—Gregorio Ferreira, data scientist and Industry 4.0 specialist

Under the hood: Intel® Distribution of OpenVINO™ toolkit

3

Intel does not control or audit third-party data. You should review this content, consult other sources, and confirm whether referenced data are accurate.Intel’s compilers may or may not optimize to the same degree for non-Intel microprocessors for optimizations that are not unique to Intel microprocessors. These optimizations include SSE2, SSE3, and SSSE3 instruction sets and other optimizations. Intel does not guarantee the availability, functionality, or effectiveness of any optimization on microprocessors not manufactured by Intel. Microprocessor-dependent optimizations in this product are intended for use with Intel microprocessors. Certain optimizations not specific to Intel microarchitecture are reserved for Intel microprocessors. Please refer to the applicable product user and reference guides for more information regarding the specific instruction sets covered by this notice. Notice Revision #20110804. Cost reduction scenarios described are intended as examples of how a given Intel-based product, in the specified circumstances and configurations, may affect future costs and provide cost savings. Circumstances will vary. Intel does not guarantee any costs or cost reduction.Intel technologies’ features and benefits depend on system configuration and may require enabled hardware, software or service activation. Performance varies depending on system configuration. No product or component can be absolutely secure. For more complete information about performance and benchmark results, visit intel.com/benchmarks.1. “The Fourth Industrial Revolution: Technology Alliances Lead the Charge,” Intel. https://www.intel.com/content/www/us/en/industrial-automation/fourth-industrial-revolution-white-paper.html.2. “The Industry 4.0 Paradox,” Deloitte, https://www2.deloitte.com/us/en/insights/focus/industry-4-0/challenges-on-path-to-digital-transformation/summary.html?id=us:2ps:3gl:confidence:

eng:cons:050519:nonem:na:47Ft6rYY:1150326353:346939503586:b:Internet_of_Things:Industry_4.0_Paradox_BMM:nb. © Intel Corporation. Intel, the Intel logo, and other Intel marks are trademarks of Intel Corporation or its subsidiaries. Other names and brands may be claimed as the property of others.1220/JC/CMD/PDF 342342-002US