Breaking the Bottleneck | Issue 25
[9/24/2023] Inventor Updates, Intel's New Substrate, UAW Strike, Uploaded Market Maps & Fred Again
Breaking the Bottleneck is a weekly manufacturing technology newsletter with perspectives, interviews, news, funding announcements, and a startup database. For an older version of a discrete and continuous manufacturing market map click the link here!
Content I Enjoyed This Week 🏭🗞️🔬
News:
What is New in Inventor 2024? [Engineering.com]
Autodesk released Inventor Professional 2024, including over 100 enhancements based on customer feedback. Some notable features and improvements include:
Design Intent: Autodesk Inventor 2024 introduces a new "Finish" feature that allows users to specify both the appearance and manufacturing details of parts. This includes material coatings, surface finishes, heat treatment, and paint. Finish settings are listed in a dedicated browser folder for easy management.
Parameter Enhancements: The software now includes a new "Hardness" unit type for use with the Finish command. Users can export both text and Boolean values from the Parameters dialog as custom iProperties for use in drawings or 3D annotations.
Mark Improvements: Building on the Mark feature introduced in Inventor 2023, users can now use non-coplanar sketches to define mark sketch geometry, wrap or project sketches to non-planar faces, and more effectively create laser marking and engraving features.
Fix, Fit, Finish: Autodesk has made various enhancements to improve workflows, including color scheme settings for drawing sketch elements, GPU ray tracing support, and the ability to cancel sketches for all 2D and 3D sketch workflows.
Section Views: Section view definitions now persist in the model design view representation, eliminating the need to recreate section views when adjusting settings. The user interface for section views has also been streamlined for efficiency.
Component Patterning Enhancements: Inventor 2024 offers improved component patterning, allowing users to select revolved, cylindrical, or conical faces to define the direction of patterns. It also introduces a new Circular Pattern Positioning Method for more control over component patterns.
Tube and Pipe: The software now includes a "Custom Elbows" option for rigid pipe runs, allowing users to add elbows of any angle. This feature enhances flexibility when working with pipe designs.
Derive Enhancements: The Derive process has been improved, offering better control over component bounding boxes, envelope size adjustments, and the creation of read-only parameters in simplified files.
Interoperability: Autodesk continues to focus on improving data exchange with other software products. Users can now send a part directly to Fusion 360 for manual inspection tasks and receive prompts to update Fusion 360 models when they are out of sync with Inventor models.
Documentation and Communication: Inventor 2024 introduces Revision Clouds as a full-fledged object and feature. These closed cloud-shaped things, commonly used for marking up drawings, are now integrated into the software. Other improvements include the ability to delete Origin Indicators, enhanced edge symbols based on ISO standards, and advancements in Model-Based Definition (MBD).
Why Apple’s 2nd Gen UWB Chip is Exciting? [Estimote]
The second-generation Ultra Wideband (UWB) chip in the newly announced iPhone 15 and Apple Watch Series 9 offers a step-change improvement in power consumption, spatial awareness, payments, and radar and health monitoring. Manufactured using a smaller 7nm process, it offers improved user experiences and longer battery life for UWB-enabled devices like AirTags, potentially extending their lifespan to 2-3 years on a single coin battery. Furthermore, the lower power UWB chip could support techniques like Downlink TDoA (time difference of arrival), enabling passive receipt of signals from UWB beacons for precise indoor positioning and radar monitoring for counting people in a room and detecting human heartbeats. Moving forward the UWB chip may interact with Apple's smart glasses (Vision Pro), enabling more power-efficient spatial orientation within a room and potentially leading to lighter and more compact smart glasses in the future. It represents an exciting development in spatial and location technologies.
Intel Unleashes Glass Substrate to Improve Compute [Intel]
Intel has announced one of the industry's first glass substrates for advanced packaging, which is expected to be used in the latter part of this decade. Compared to organic substrates, glass offers benefits such as ultra-low flatness, better thermal and mechanical stability, and higher interconnect density. This advancement will allow chip architects to create high-density, high-performance chip packages for data-intensive workloads, including artificial intelligence (AI). Intel plans to deliver complete glass substrate solutions to the market in the second half of this decade, helping to advance Moore's Law beyond 2030. Glass substrates are seen as a necessary step for the next generation of semiconductors, as the industry approaches its limits for scaling transistors on silicon packages using organic materials. Glass substrates will be initially used in applications requiring larger form factor packages, such as data centers and AI.
Toyota Develops New AI for Training Robots [New Atlas]
Toyota, in collaboration with MIT and Columbia Engineering, has developed an innovative learning approach called Diffusion Policy that enables robots to quickly learn and adapt to various tasks based on human instructions or demonstrations. Unlike some startups that use VR telepresence to teach robots, Toyota's approach focuses on haptics, providing operators with tactile feedback through hand controls. After a human operator demonstrates a task multiple times, the robot's AI constructs its own internal model of success and failure, running numerous simulations to refine its techniques. Toyota aims to have hundreds of tasks mastered by the end of the year, with a goal of over 1,000 tasks by the end of 2024. They are developing what they call a Large Behavior Model (LBM), a comprehensive framework that will allow robots to rapidly learn new skills. As the LBM evolves, it is expected to enable humanoid robots to control a wider range of tools designed for human use. The technology's ability to adapt to deformable objects, cloth, and liquids makes it particularly promising for a wide range of applications.
Wonking Out: Making Manufacturing Good Again [NYTimes]
Given the United Auto Workers’ (UAW) efforts to seek higher wages and benefits in their negotiations with automakers, the Op-ed evaluates the request in the historical context of manufacturing jobs in the YS. Manufacturing jobs weren't inherently good before the New Deal, with many workers receiving low wages and working long hours. However, the 1930s and 1940s saw a significant rise in real manufacturing wages, mainly due to unionization and the favorable political environment of the New Deal. This period, known as the "Great Compression," reduced wage inequality and created a middle-class society. However, real wages in manufacturing have been declining since 1980, especially in auto manufacturing. The decline is attributed to the decline in unionization rates and union power since their peak in the New Deal era. The decline in unionization has eroded the wage premium that manufacturing jobs once enjoyed. Consequently, the piece argues that the restoration of worker bargaining power will be found through unions and that this should ideally spread to other sectors to improve worker conditions.
NVIDIA and Mercedes-Benz Partner for Factory Digital Twin [VentureBeat]
Mercedes-Benz is partnering with Nvidia to create digital twins of its real-life factories using Nvidia's Omniverse platform. These digital twins will be metaverse-like representations of car factories and will enable Mercedes-Benz to optimize its manufacturing processes, reduce costs, and collect real-life data for future designs. The digital twins will be used in production applications for designing, collaborating, planning, and operating manufacturing and assembly facilities. The initiative aims to enhance efficiency, avoid defects, and save time, leading to greater flexibility, resilience, and intelligence in Mercedes-Benz's production system. The digital twins will be implemented in factories in Germany, Hungary, and China, offering a blueprint for over 30 Mercedes-Benz factories worldwide. This approach will facilitate the production of electric, hybrid, and gas models on the same assembly lines and enable vehicles to roll off the production line with the latest software. Additionally, the digital twin system will help reduce energy consumption and costs. This partnership is part of Mercedes-Benz's broader effort to adopt AI, digital twins, and digital production to meet the demand for electric, hybrid, and combustion car models on a single assembly line while meeting sustainability goals.
Research:
World Fab Forecast Report [Semi]
Some highlights:
Global fab equipment spending for front-end facilities in 2023 is expected to decline 15 percent year-over-year (YoY) to US$84 billion from a record high of US$99.5 billion in 2022 before rebounding 15 percent YoY to US$97 billion in 2024
Memory spending in 2024 is forecast to climb 65 percent to US$27 billion after a 46 percent decline in 2023.
DRAM investments are expected to decline 19 percent YoY to US$11 billion in 2023 but recover to US$15 billion, a 40 percent annual jump, in 2024.
NAND spending is projected to mirror that trend, decreasing 67 percent to US$6 billion in 2023 but surging 113 percent to US$12.1 billion in 2024.
MPU investments are expected to remain flat in 2023 and increase 16 percent to US$9 billion in 2024.
Podcasts:
The Industrial Waste: Manufacturing Misguided Policies [Economist]
The Impact of AI in Manufacturing [Supply Chain Now]
Lumafield Voyager [Lumafield]
Chart of the Week:
The U.S. Lost 4.1 Million Days of Work Last Month to Strikes Even Before the UAW Strike
Track the Clean Manufacturing Factories Being Built Across the US
Manufacturing Deals
Dragos - An industrial cybersecurity company providing companies with visibility into your ICS/OT assets, vulnerabilities, threats, and response actions, and supports you with forensics and OT-scanning.
$74 million [Series D Extension] - Led by WestCap
Jura Bio - A company using using machine learning and synthetic biology to develop new therapeutics
$16.1 million [Venture] - Including Michael Chambers, John Ballantyne, Fontus Capital, Josh Elkington and others
Weekly Planned Downtime
Fred Again Actual Life 3 with Zane Lowe