🎯Autowist Report 96

GM Ultifi and Carplay, Xpeng AI Day, BMW Q3 Presentation, Tesla Vision Patent, Thermal Imaging for Autonomous Driving

Author

Harikrishnan Radhakrishnan
November 07, 2024

IN TODAY’S REPORT

💼BRIEFING

Top insights and analysis that is moving the needle in the automotive industry

Why GM’s Software Boss Thinks Ultifi Can Beat Apple CarPlay (Link)

GM's software chief, Baris Centinok, believes the automaker’s new in-house Ultifi system can provide a more integrated and seamless experience for drivers than Apple CarPlay or Android Auto.

Having previously worked at Apple, Centinok argues that GM’s decision to drop these popular smartphone mirroring options allows for deeper integration, especially as GM vehicles move toward semi-autonomous driving capabilities. There is a small pitfall here. GM does not enjoy the same market penetration and ecosystem that Apple had with the iPhone. At the same time, I understand that GM is strategizing towards that.

Ultifi aims to streamline functions such as navigation, entertainment, and advanced features like Super Cruise, aligning with the car's internal systems in ways that smartphone mirroring can't. This is one of the ways in which I think that over time OEM based infotainment systems can beat CarPlay.

Centinok acknowledges that some see this shift as a profit-driven move to keep subscription revenue within GM, but he insists the primary motivation is delivering a better customer experience. He envisions Ultifi adding value through services and optional entertainment features that align with GM's vision of connected, tech-enhanced vehicles.

XPENG AI Tech Day Highlights

  1. The company unveiled a proprietary AI chip, the Turing chip, designed to power vehicles, robots, and flying cars, supporting models with up to 30 billion parameters.

  2. The Turing chip and new "Canghai Platform" aim to support full Level 4 autonomous driving, promising robotaxis by 2025.

  3. XPeng’s AeroHT division updated its flying car development, with public demonstrations of its Land Air Carrier scheduled soon and pre-orders opening in December.

  4. XPeng introduced the Kunpeng Super Electric System, a gas-powered range extender for longer trips.

  5. This hybrid system offers a total range of 870 miles with ultra-fast charging capabilities on an 800V platform, allowing an 80% charge in 12 minutes.

BMW Q3 Investor Presentation

Rivian’s software boss is ‘super excited’ about third-party apps (Link)

Rivian’s Chief Software Officer, Wassym Bensaid, discussed his vision for Rivian’s software future at TechCrunch Disrupt 2024. He emphasized Rivian’s unique approach to car software, noting the company’s preference for a button-free, digital interface and its decision to forgo phone mirroring systems like CarPlay. Instead, Rivian aims to build its own app ecosystem by offering APIs for third-party developers to create custom applications compatible with Rivian’s vehicles, potentially creating new revenue streams.

Rivian has already collaborated with TezLab, a company known for its Tesla-compatible app that tracks vehicle statistics. Bensaid shared that Rivian is engaging with other developers and is working to complete a more robust software development kit (SDK) to support this ecosystem, a project he expects to complete within a few quarters. Bensaid also emphasized his belief in collaborating with the broader tech ecosystem, supporting open-source development, and leveraging external feedback to enhance Rivian’s platform.

⚡ROUNDUP

Key headlines shaping the auto industry this week

  1. VW CEO Admits “Decades of Structural Problems” (Link)

  2. Volvo Cars to acquire full ownership of NOVO Energy JV (Link)

  3. NV300 Max Thermal Imaging Camera Promises To Be the Difference Between Life and Death (Link)

  4. Lucid (LCID) is gaining momentum in Europe’s largest auto market with its luxury EVs (Link)

  5. Inside BMW's Next-Generation EV And Battery Plant In Mexico (Link)

  6. Rivian’s software boss is ‘super excited’ about third-party apps (Link)

  7. The Complete Guide To Electric Vehicle Platforms (Link)

  8. Mazda Recalls CX-70 and CX-90 Sport Utility Vehicles for Software Issue (Link)

  9. Why GM’s Software Boss Thinks Ultifi Can Beat Apple CarPlay (Link)

  10. Nio Registrations in Germany Drop to 15 in October, Lowest Since February 2023 (Link)

  11. Scout's South Carolina Plant Could Build EVs For Volkswagen And Others (Link)

  12. New Aston Martin boss outlines plans for profitability (Link)

  13. Daimler Truck subsidiary Torc Robotics achieves driver-out validation milestone (Link)

  14. Elektrobit's EB corbos Hypervisor achieves ISO 26262 ASIL-B safety certification (Link)

  15. Ford enhances EV navigation with Google Maps integration (Link)

  16. Tesla Raises Wages at German Gigafactory Following Conversion of 500 Positions (Link)

  17. Toyota in $3.3Bn JV to Bring Autonomous Driving to Roads by 2030 (Link)

  18. FAW-Volkswagen's all-new Tayron L goes into series production (Link)

  19. Mercedes: electric car battery recycling won't stop mining (Link)

  20. BYD Added A Tesla-Worth Of Production Capacity Over The Past 3 Months … With More To Come (Link)

  21. Tesla 2024 Holiday Update Starts Testing With Supercharger and Cold-Weather Improvements (Link)

  22. 1 in every 2.8 electrified cars sold in China is a BYD (Link)

  23. MIPI Alliance releases security framework for image sensor data protection (Link)

🔬PATENTS

Fresh Innovations from global automotive OEMs

(Tesla) VISION-BASED SYSTEM TRAINING WITH SIMULATED CONTENT - US20240355132A1

The patent discusses a system to manage and process vision data for vehicles, using machine learning algorithms and ground truth labels to enhance autonomous vehicle operations. Vision-only systems process data from cameras to identify objects and their attributes, such as position, velocity, and acceleration.

System Components:

  • Vision System: Uses vehicle-mounted cameras to capture images for object detection.

  • Network Service: A network-based service collects, processes, and generates ground truth labels for training data.

  • Machine Learning Algorithms: Algorithms are trained with labeled data to recognize and track objects without relying on additional sensors like LIDAR or RADAR.

Process:

  • Data is captured from cameras at different points in time.

  • Ground truth labels (e.g., object position or velocity) are created for each data capture instance.

  • Later captures are used to verify and refine labels for earlier data, improving accuracy over time.

System Architecture:

  • Vision Information Processing: Processes captured image data and labels, updating algorithms for accurate real-time object detection.

  • Local and Network Storage: Stores labeled data and processed information for future training and real-time use.

Claims:

  • The system can process time-sequenced image data to refine object attributes like yaw, velocity, and position.

  • Supports dynamic capture rates for enhanced object tracking (e.g., 24 Hz).

 Application: Suitable for vision-only autonomous systems, enabling enhanced object tracking and position estimation for self-driving applications.

🤖TECH DEEP DIVE

An inside look into new automotive technology

NV300 Max Thermal Imaging Camera Promises To Be the Difference Between Life and Death (Link)

The NV300 Max is a thermal imaging dashcam developed by Thermal Master to enhance driver safety, particularly in low-visibility conditions. It uses a thermal camera to detect heat signatures up to 300 yards away, making it ideal for spotting animals, pedestrians, and vehicles in dark or foggy environments. The system includes a car-mounted camera and a dashboard display that provides live thermal feeds. Additionally, the NV300 Max offers audible and visual alerts, operates in extreme temperatures, and can function as a diagnostic tool by monitoring vehicle health via smart device connectivity.

Technical Breakdown:

1. Thermal Imaging Camera:

  • Resolution: 512x384 thermal sensor that captures heat signatures.

  • Detection Range: Capable of identifying objects up to 300 yards away.

  • True Color Imaging: Generates both thermal and color images.

  • All-Weather Operation: Works in rain, fog, snow, and extreme temperatures (-20°C to 80°C).

  • Durability: The device is dust-proof, water-resistant, and equipped with defogging and defrosting capabilities.

2. Display System:

  • Tablet-like Display: Shows live thermal feed and color images.

  • Mounting: Magnetic and telescopic mount for easy dashboard placement.

  • Alerts: Provides visual and auditory alerts every 0.1 seconds when a potential hazard is detected.

3. Diagnostic Capabilities:

  • Vehicle Monitoring: Tracks engine performance, battery status, and error codes.

  • Connectivity: Supports Wi-Fi and Bluetooth for integration with smartphones via an app.

4. Installation:

  • Requires vehicle wiring for setup but enhances safety with early detection technology, reducing accident risks in low-visibility scenarios.

This system is available on Kickstarter with an early bird price around $400, making it relatively accessible for those seeking enhanced vehicle safety features.

🐦AUTO BUZZ

Top posts, tweets, and videos from the automotive industry

📈MARKET SNAPSHOT

This week’s key movers in the automotive stock market

📆CALENDAR

Upcoming auto industry events and and earnings calls