MEDIA TEK IoT Innovation Industrial Manufacturing User Guide
- September 28, 2024
- MEDIA TEK
Table of Contents
IoT Innovation Industrial Manufacturing
“`html
Product Specifications
-
Product Name: IoT Innovation Guide: AI and the Future of
Industrial Manufacturing -
Manufacturer: Not specified
-
Product Type: Information guide/manual
Product Usage Instructions
Vision Systems
Vision systems like cameras or LIDAR/radar paired with AI and
machine learning can be used for various industrial
applications:
-
Augmented Workforce Efficiency: Utilize robots
or cobots to handle repetitive or dangerous tasks, improving
manufacturing capacity. -
Quality Assurance (QA): Implement automated
optical inspection (AOI) to analyze images from the production line
for product quality. -
Process Improvement: Analyze visual data to
identify and resolve machine/process bottlenecks, reducing downtime
and suggesting improvements. -
Access Management: Use cameras and edge visual
processing for tasks like facial recognition, clocking in/out, and
access control.
Speech-to-Text Systems
Voice recognition systems can be beneficial for industrial
settings:
- Utilize always-on microphones connected to an SoC to detect
words and sounds for understanding human language.
Frequently Asked Questions (FAQ)
Q: How can I apply AI to the production ecosystem?
A: To apply AI in industrial manufacturing, consider use cases
like plant consumption management, quality sensing, smart
conveyance, factory asset intelligence, and more. Utilize
technologies such as vision systems, speech-to-text systems, and
enhanced connectivity for efficiency and productivity.
Q: What are the benefits of using vision systems in
manufacturing?
A: Vision systems help enhance workforce efficiency by
automating tasks, improve quality assurance through automated
inspection, identify process bottlenecks for improvement, and
manage access control using facial recognition.
“`
IoT Innovation Guide: AI and the Future of Industrial Manufacturing
Introduction
The use of AI poses questions about what the future of work will look like,
among them: “What are the possibilities for smart, connected industrial
manufacturing systems?” AI extends the capabilities of machines to sense,
learn, reason, and interact with humans in ways that are natural, intuitive,
and complementary to the way we work. By pairing AI with IoT (Internet of
Things) and connecting insights across the production process, manufacturers
can achieve higher quality final products, while also increasing the safety,
efficiency, and sustainability of industrial processes. In fact, a recent
McKinsey survey found that AI has the potential to unlock approximately $1
trillion in value from the global industrial sector.1
1 Source: https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-
state-of-ai-in-2023generative-ais-breakout-year
| 2 IoT Innovation Guide ® MediaTek 2024
How to Apply AI to the Production Ecosystem
How can industrial manufacturers capture value from AI? A recent survey by Deloitte identified 8 use cases for smart manufacturing:2
Use Cases for Smart Manufacturing
Some of the innovative IoT and AI technologies behind the top use cases include:
icon
icon
Plant consumption and energy management
Quality sensing and detecting
Smart Conveyance
Factory asset intelligence– performance management
Vision Systems Speech-to-Text Systems Enhanced Connectivity Smart Power Consumption
INTRODUCTION
icon
icon
icon
icon
Digital product development–
digital twin
Factory synchronization
and dynamic scheduling
Augmented workforce efficiency
Additive manufacturing
Read on to learn how to use these technologies in your industrial processes.
2 Source: https://www2.deloitte.com/us/en/insights/industry/manufacturing/
accelerating-smart-manufacturing.html
3 |
Vision Systems
Vision systems like cameras or LIDAR/radar provide powerful sensing and
surveillance that, when paired with AI, machine learning and connectivity, can
recognize faces, spaces, products, parts, and scenes, even in low-light
conditions. To make the algorithm faster and capable of real-time interactions
with other systems or humans, use of high-speed edge processing and visual
Simultaneous Localization and Mapping (SLAM) is required.
| 4 IoT Innovation Guide ® MediaTek 2024
VISION SYSTEMS
Industrial Applications
Augmented workforce efficiency: Let robots (or collaborative robots, known as
`cobots’) take on repetitive or dangerous aspects of the production process,
while adding manufacturing capacity with 24/7 availability.
Quality assurance (QA): Use automated optical inspection (AOI) to unlock big
data insights that can help identify product quality using images from the
production line.
The MediaTek Advantage at the Edge
MediaTek Genio supports high-speed camera processing and visual SLAM, adding
computer power with GPU & AI engines right from the highly capable System on a
Chip (SoC).
5 |
Process improvement: Identify and correct machine or process bottlenecks using
visual data analysis to help save costs from unnecessary downtime and
recommend process improvements.
Access management: Use cameras and edge visual processing for facial
recognition, clocking in/out on the manufacturing floor, signing in to
industrial PCs, or providing access control to locks and doors within a
building.
Did you know? A joint study by Vanson Bourne and GE found that
23%
of all manufacturing downtime was caused by human error, which if corrected by
machine support, could generate substantial savings.3
3 Source: https://lp.servicemax.com/rs/020-PCR-876/images/After%20The%20
Fall%20whitepaper%20-%20updated%20global%20numbers%20FINAL%20 refresh.pdf
Speech-to-Text Systems
Voice recognition uses always-on microphones connected to an SoC that runs
software to help machines detect words and sounds to understand human language
as it is spoken, heard from a distance and isolated from noise.
With New Language Processing (NLP), vast amounts of audio data can be
processed without even having to send it to the cloud, saving on cloud storage
and processing costs, and reducing concerns about data privacy and
confidentiality by keeping this data at the edge, in the device. With speech-
to-text systems, workers can give commands, ask questions, and record
observations and issues instantly, without having to stop handling the
machine, or removing protective gear.
| 6 IoT Innovation Guide ® MediaTek 2024
Industrial Applications
Voice commands: Talk to the industrial machine or robot like a voice
assistant; for example, ask it to retrieve data from the backend manufacturing
system instantly without having to stop handling the machine or remove
protective gear.
Ask for instructions: Ask for instructions, hands-free, for guidance or help
with training and safety.
Take notes: Ask the machine to keep track of key observations or relay the
current situation.
The MediaTek Advantage at the Edge
Genio enables New Language Processing (NLP) on the device, which eliminates
the need for a cloud-supported dictionary for audio language processing (ALP).
7 |
SPEECH-TO-TEXT SYSTEMS
Enhanced Connectivity
For AI to support manufacturing systems and workers in real time, IoT
connectivity must have minimal latency to analyze data and make split-second
decisions, where even a second or two of latency from sending data to the
cloud and waiting for a decision could have business consequences. Enhanced
processing at the edge helps to process large data volumes closer to the
source in order to reduce latency to the cloud.
With edge AI, processing happens where the data is created, instead of being
sent offsite to a remote data center. Industrial gateways process data
efficiently and safely on the manufacturing floor with an intelligent hub that
interfaces between local networks and the cloud.
| 8 IoT Innovation Guide ® MediaTek 2024
Industrial Applications
Real-time interactions: React to visual data and voice commands with minimized
latency.
Split-second decision making: Process data quickly to prevent delays in
decision making, especially amid dangerous activities.
Over-the-air updates: Add software features with non-disruptive over-the-air
updates.
EDGE AI
ENHANCED CONNECTIVITY
The MediaTek Advantage at the Edge
Genio brings connectivity anywhere with a 5G broadband or 5G RedCap connection
that keeps manufacturing systems in contact, even when large amounts of data
need to be transferred.
9 |
Smart Power Consumption
In smart systems that require vision or voice recognition, sensors (like
microphones) are always on, requiring power-saving strategies, especially for
batterypowered devices. Traditionally, in complex functions like facial
recognition, the CPU can account for more than 40% of power consumption,
contributing to the 1% of all global GHG emissions caused by end-user
computing. Using dedicated AI processors can produce significant power-
savings, as AI algorithms can be used within the design of the SoC hardware
and software to help reduce and optimize power consumption.
| 10 IoT Innovation Guide ® MediaTek 2024
Industrial Applications
Energy efficiency: Operate more sustainably, reduce the carbon footprint and
cost of your data centers.
Maintenance reduction: Reduce costly maintenance of long-term installations
and remote robots that require connectivity system upgrades or replacements
due to diminished battery power.
Always-on connectivity: Stay connected throughout critical manufacturing
processes and reduce downtime caused by power failures.
The MediaTek Advantage at the Edge
To reduce power consumption, MediaTek engineers have pioneered several system
optimizations.
· MediaTek’s hardware cuts power consumption of internal components, such as the CPU, GPU, and AI processor when they are not being used, optimizing the operating power consumption in response to daily demands.
· MediaTek’s Software Development Kits, such as the NeuroPilot SDK, allow developers to adapt and optimize their software for MediaTek AI processors, taking advantage of faster, more efficient processing.
SMART POWER CONSUMPTION
MediaTek NeuroPilot automatically works with all MediaTek Genio platforms and
beyond, creating a multi-purpose software that helps reduce development costs
and time to market.
11 |
Use Cases for Smart Manufacturing
Reduce development costs and time to market
Implement MediaTek Genio in Your Industrial IoT Manufacturing Processes
MediaTek provides a suite of advanced, intelligent IoT SoCs and an ecosystem
of partners and smart modules. Together, we enable industrial manufacturers to
develop intelligent devices for highly demanding, edge-weighted applications
that need heavy IO, reliable connectivity, and the immediacy and reliability
of embedded processing.
Designed for demanding AI and performance-centric IoT applications, MediaTek
Genio is a premium IoT SoC that empowers the edge with multiple in-chip
processors and extremely capable multitasking performance in the latest Open
OS.
· Best-in-class CPU · Advanced 3D Graphics · Powerful AI performance ·
Hardware support for the
latest multimedia standards
· Multiple 4K displays · Exceptional power efficiency · Industrial-grade longevity
Discover MediaTek Genio
IoT Innovation Guide ® MediaTek 2024
References
Read User Manual Online (PDF format)
Read User Manual Online (PDF format) >>