Breaking the Bottleneck | Issue 76
[3/24/2024] NVIDIA, MHI Annual Industry Report, Robots, & More!
Breaking the Bottleneck is a weekly manufacturing technology newsletter with perspectives, interviews, news, funding announcements, manufacturing market maps, and a startup database!
💥 If you are building, operating, or investing in manufacturing, hardware, or robots, please reach out. My email is aditya@schematicventures.com – I’d love to chat!
🏭 If you were forwarded this and found it interesting, please sign up!
Content I Enjoyed Last Week 🗞️🔬 📚
News:
Newton, an Open-Source Physics Engine for Robotics Simulation [NVIDIA]
Physical AI models enable robots to autonomously perceive, interpret, reason, and interact effectively with the real world. To address the longstanding “sim-to-real” gap, NVIDIA, Google DeepMind, and Disney Research have introduced Newton, an open-source, extensible physics engine built on NVIDIA Warp, an acceleration library leveraging NVIDIA GPUs. Newton supports advanced robotic simulations, allowing developers to accurately model diverse physics-based behaviors, from rigid and soft-body dynamics to complex contact interactions. Notably, it integrates seamlessly with robotics frameworks such as MuJoCo Playground and NVIDIA Isaac Lab, enabling significant performance improvements like a 70x acceleration in humanoid simulations and 100x in manipulation tasks. Newton’s architecture emphasizes differentiable gradients which enhances learning and optimization capabilities. The platform’s extensibility also supports custom solvers, complex multi-physics scenarios (like interaction with deformable materials), and a unified data structure based on OpenUSD, streamlining robotic workflows and interoperability. Disney Research is pioneering Newton’s use in entertainment robotics, demonstrated through its expressive Star Wars-inspired BDX droids, and is collaborating to define standardized OpenUSD robotic data structures. This initiative positions Newton as a critical step forward in bridging simulation and real-world robotic performance, benefiting research, industry, and next-generation robotic applications.
From Ford to Musk, the Perils of Trying to Build a Global Auto Empire [WSJ]
The ambition to build global automotive empires, dating back to Henry Ford’s international expansion, faces increasing hurdles today as Tesla, initially successful with a global single-brand strategy, grapples with mounting challenges. Tesla’s stock recently halved due to consumer backlash against CEO Elon Musk’s political involvement, particularly affecting European sales, and intensifying protectionism, notably from the Trump administration’s tariff threats. Simultaneously, stark regional differences in electric vehicle (EV) adoption rates complicate global market strategies, making uniform global offerings increasingly difficult. As a result, regional automakers like France’s Renault have emerged stronger by leveraging partnerships, localized supply chains, and targeted budget-friendly EVs, reflecting investors’ shifting confidence toward smaller, agile players. In contrast, conglomerates such as Stellantis face difficulties managing multiple brands, leading to inefficiencies and stagnant growth. Even Toyota and Volkswagen, traditional benchmarks of successful global automotive strategies, face obstacles catering simultaneously to diverse regional preferences and varying EV adoption rates, indicating that the industry’s electrification trend might ultimately favor regional specialization over global standardization.
Figure AI unveils BotQ Manufacturing Facility [Figure]
Figure AI has unveiled BotQ, a high-volume factory designed to manufacture up to 12,000 humanoid robots annually, bringing its production capabilities in-house to improve quality, scalability, and cost efficiency. The company has re-engineered its robot architecture by reducing part count, switching from CNC machining to high-throughput methods like injection molding and diecasting, and establishing a robust MES/ERP/PLM/WMS software foundation. It has also formed specialized safety and reliability teams to test lifecycle performance and drive design improvements. With Figure 03, its next-gen production robot, the company is prioritizing manufacturability and affordability while preparing for high-volume output. Facing a lack of established supply chains for humanoids, Figure has designed critical components such as actuators, motors, batteries, and electronics in-house and partnered with vendors capable of scaling to 100,000 robots. It has hired global supply managers and manufacturing engineers to design flexible production lines, balancing manual labor with selective automation. For example, automated grease stations and battery cell testers increase speed and quality. By integrating its MES with IoT devices and deploying its robots for material handling and assembly, Figure is building a hybrid workforce and positioning itself to lead in scalable humanoid manufacturing.
Isaac GrootN1 [NVIDIA]
Powering the Remanufacturing Renaissance with AI [McKinsey]
Companies across sectors are increasingly turning to remanufacturing to address supply chain shortages, reach new customer segments, and create high-margin revenue streams. However, they face challenges such as forecasting the availability of returned parts (core) and pricing complex, low-volume SKUs. AI technologies are now helping remanufacturers tackle these obstacles by enhancing core forecasting accuracy, improving SKU pricing precision, and streamlining warranty claims management. For instance, specialized AI forecasting tools can analyze historical part usage and macroeconomic trends to reduce safety stock by 2-4% and freight costs by 3-5%, while AI-driven pricing optimization has proven to boost margins by 2-4%, and one case, up to 11-15%. Additionally, generative AI and large language models are transforming warranty management by rapidly extracting insights from unstructured text data, enabling quicker fault interventions and substantially reducing warranty-related costs by 5-10%. Successful implementation of AI in remanufacturing extends beyond just technology. Companies must prioritize focused use cases, secure strong executive sponsorship, ensure strategic alignment, assess organizational capabilities thoroughly, plan clear implementation road maps, and invest heavily in change management to realize the full benefits of AI-driven transformation.
Medtech Can Improve Quality and Regulatory Processes with GenAI [BCG]
Medtech companies have been slow to adopt generative AI (GenAI), with only 10% seeing measurable benefits, significantly behind other sectors such as biopharma, which has already successfully integrated GenAI into processes like regulatory filings and quality management. Regulatory and quality processes in MedTech present major bottlenecks due to complex documentation requirements, making them ideal candidates for GenAI applications. Early adopters have seen substantial efficiency gains: GenAI-generated first drafts have reduced writing times by up to 60-70% in regulatory labeling, clinical trial protocols, and clinical study reports while improving accuracy and consistency across documents. The most promising medtech applications of GenAI include automating technical documentation writing, streamlining medical, legal, and regulatory (MLR) reviews, drafting product manuals, managing complaint handling, creating quality documents, and improving deviation management. GenAI can automate initial drafts, analyze data to ensure regulatory compliance, enhance collaboration among cross-functional teams, and manage interdependencies across multiple documents. Companies can leverage existing GenAI solutions (deploy), customize and integrate tools to reshape entire processes (reshape) or develop entirely new solutions (invent), though most medtech firms will find significant benefits from customizing existing platforms. Ultimately, implementing GenAI effectively requires structured processes, clear human oversight, and customized integration to realize significant productivity and accuracy improvements in regulatory and quality management tasks.
Boston Dynamics Atlas
Dexterity Mech [Dexterity]
Research:
2025 MHI Annual Industry Report [MHI & Deloitte]
The MHI 2025 Annual Industry Report emphasizes the critical importance of end-to-end (E2E) supply chain orchestration as organizations face increasingly complex challenges, from economic uncertainty and inflation to workforce shortages. Surveying over 700 global supply chain leaders, the report highlights inflation (38%), economic uncertainty (37%), workforce and talent shortages (35%), agility and resiliency (28%), and inventory management (25%) as the most impactful supply chain trends currently shaping the industry. Other interesting takeaways:
Technology adoption is set to rise dramatically over the next five years, especially in areas like Inventory and Network Optimization (projected to reach 92% adoption from 66% today), Cloud Computing & Storage (91% from 58%), Predictive Analytics (87% from 49%), Robotics & Automation (83% from 41%), Artificial Intelligence (82% from 28%), Sensors & Automatic Identification (88% from 66%), Internet of Things (77% from 40%), and Autonomous Vehicles & Drones (64% from 15%).
Despite this optimistic outlook, companies still encounter barriers to adopting these technologies, including a lack of budget (26%), limited understanding of technology (22%), and unclear business cases (19%), particularly for AI implementation.
Currently, the main uses of AI in supply chains include inventory management (35%), demand forecasting and warehouse management (34%), and logistics and transportation (27%).
Addressing talent shortages, the report reveals that 63% of organizations are actively upskilling current employees, while 36% are recruiting specifically for different skill sets. Although only 19% have fully embraced a skills-based approach, organizations that do report that their employees are 52% more likely to innovate. The report cites IKEA’s AI bot, “Billie,” as an example of effectively combining AI automation with strategic employee upskilling, allowing humans to focus on creative, higher-value tasks.
Mapping 4000 Global Industrial Automation Projects [IoT Analytics]
Podcasts/Video:
The Great Jones Act Debate [Odd Lots]
Finance & Transactions 💵
Venture Capital:
Apptronik - A company building humanoid robots.
$53 million [Series A2] - From new investors Mercedes-Benz, Japan Post Capital, ARK Invest, RyderVentures, and others.
Augment - A company building autonomous AI assistants for freight and trucking brokerage.
$25 million [Seed] - Led by 8VC
Rerun - A company building a multimodal data stack for physical AI
$17 million [Series A] - Led by Point Nine and joined by Sunflower Capital and insiders Costanoa Ventures and Seedcamp.
Tera AI - A company building a spatial reasoning AI system to provide affordable visual navigation for autonomous robots.
$7.8 million [Seed] - Led by Felicis and joined by Inovia, Caltech + Wilson Hill, and Naval Ravikant.
Aletiq - A company building product lifecycle management softtware for industrial manufacturers.
€6 million [Seed] - Led by Point Nine
Planned Downtime 🧑🔧
Happy Gilmore 2