The stack is primarily based on deep learning AI processing for all of its modular components, including environmental perception, sensor fusion and vehicle path planning. It was designed in conjunction with the company’s CV3-AD AI domain controller SoC to ensure it runs optimally on the SoC’s CVflow AI engines – minimizing power consumption and the processing load for its on-chip Arm cores. The SoCs also provide optimal processing for vision and radar sensing and fusion, while Ambarella’s stack more broadly offers a flexible implementation model, enabling OEMs to leverage any or all of its modules in combination with their own software.

A key feature of the AI-based software stack highlighted by Ambarella is its deep-learning-based planner, a dynamic vehicle path planning module that enables a more natural driving experience. In developing and refining this functionality, Ambarella leveraged its evolving R&D vehicle fleet which allowed the deep planner module to be trained with data based on a wide variety of real-world scenarios. The latest vehicles in this fleet are equipped with a sensing suite consisting of mono and stereo cameras, as well as Ambarella’s Oculii 4D imaging radar, with the CV3-AD processing the data sent from these sensors. While highlighting this configuration, Ambarella noted that its solutions are flexible to support OEM’s specific sensing suites (including the option for adding LiDAR).

Another feature of the stack highlighted by Ambarella was its requirement of standard-definition maps, eliminating the need for pre-generated HD maps. Here, HD maps are generated in real time using live environmental data fed from the sensing suite. This capability offers support in handling difficult AD scenarios, such as downtown areas in large cities with roundabouts, narrow roads with parked vehicles, heavy traffic, construction, and a high density of pedestrians, cyclists and other vulnerable road users.

The CV3-AD SoCs run Ambarella’s entire AD stack with enhanced power consumption, AI performance per watt, and simplified thermal management. The company said that, because its software and hardware are designed to work optimally together, OEMs can sufficiently integrate additional software to create differentiated features, including those for functional safety.

Leveraging the funds, DiDi will boost its research and development efforts, accelerate the implementation of related products, pursue open collaborations in the industry chain, develop a sustainable and open industry ecosystem, and expedite the widespread commercial use of autonomous driving technology.

Today, the autonomous driving company operates robotaxi services in certain areas of Guangzhou and Shanghai using a mixed dispatching model, which entails a mix of both autonomous vehicles and vehicles driven by humans.

Priced between RMB 3,000 ($411) to RMB 3,999 ($547), the HP170 enables autonomous driving on highways and expressways, automatic emergency braking, and short-distance memory parking. These functions are enabled by its sensor system, consisting of 12 ultrasonic radars, four fisheye cameras, two rear corner radars (as optional), one forward facing camera, and a forward-facing radar as well as 5 TOPS of computing power.

The HP370, priced between RMB 5,000 ($684) to RMB 5,999 ($821), offers 32 TOPS of computing power, two forward facing cameras, two side-view cameras, one rear-view camera, four fisheye cameras, one front radar, two rear corner radars, 12 ultrasonic radars and two front corner radars as optional. It supports additional functions, such as memory driving within a certain city, and comes with intelligent barrier avoidance.

At the top of the range is the HP570, priced between RMB 8,000 ($1,095) and RMB 8,999 ($1,232), which offers customers the choice between a chip capable of 72 TOPS, and one capable of 100 TOPS. Its sensor suite consists of two forward facing cameras, four side-view cameras, one rear-view camera, four fisheye cameras, one front radar, 12 ultrasonic radars, and optionally allows for LiDAR sensor to be added. The HP570 enables automatic driving on urban roads and is equipped with an all-scenario parking assist.

Operating as a response to the challenges faced in procuring affordable EVs certified for L4 autonomous services to meet high industry demand, Tier IV partnered with several OEMs to develop a scalable E/E architecture, and a redundant drive-by-wire module, that will enable customers to define AD features through add-on components.

Through fanfare, Tier IV will offer dedicated after-sales services – which are currently being prepared. Initially, fanfare will launch with a minibus model in 2023, before gradually expanding its vehicle lineup to include several models by 2024. The company is looking to produce 300 units of AD-enabled EVs with its partners in 2025 and will continue to develop an environment for these partners to accelerate the mass production of AD-enabled EVs for a variety of purposes.

Leveraging Tier IV’s experiences in AVs, fanfare’s EV models have been designed to accommodate various operational design domains with open specifications of AD software. The E/E architecture adopted in its vehicles, for example, meets high safety standards and features redundant designs for sensors, computers, and drive-by-wire modules dedicated to AD technology to prevent the inability to brake or steer due to a single point of failure. These design specifications are likewise standardized through international industry organizations, such as Mobility In Harmony (MIH) and the Autoware Foundation (AWF).

In its initial phase, fanfare’s baseline EVs are procured from multiple car manufacturers, upgraded to ensure compatibility with L4 AD systems, and then sold to customers as white-label EVs. Following this initial release, Tier IV will move to encourage third-party companies to mass-produce the same EV models offered through fanfare – disclosing their design specifications as set out in its L4 Custom Design Guidelines initiative. The company is also actively exploring new opportunities to engage with as many as OEMs as possible who can provide baseline EVs that will be compatible with the open specifications of fanfare’s EV models.

The system includes two version of DeepRoute.ai’s self-driving solutions for automakers – D-PRO and D-AIR. D-PRO offers HD map-free operation and features such as valet park assist and point-to-point navigation without operational design domain restrictions. D-AIR focuses on HD map-free driver assistance – packaging features such as automatic cruise control, lane centering control, and autonomous emergency braking.

While their approach may differ, DeepRoute.ai expects that both versions will help expedite the mass production plans of many OEMs as they address high mapping and maintenance costs, alongside the limitations of geo-fencing. Driver 3.0 is further capable of adaptive cruise control, stop and go, obstacle avoidance, unprotected left turns, and other technologically complex maneuvers.

In a video demonstration of Driver 3.0, the company showcased the system’s ability to generate lane-level information around the car in real-time, without HD maps. The vehicle shown drove autonomously around Shanghai during peak traffic hours, interacting with pedestrians and other road users safely and consistently across a variety of road conditions.

While full details on the new system’s rollout have not yet been detailed, DeepRoute.ai confirmed that it will be available to purchase in 2023 from an established automotive brand. Following this launch, the company confirmed that Driver 3.0 will benefit from over-the-air software updates – providing more improved functions and services over time.

The system’s new generation offers improvements in areas such as speed, tele-assist capabilities, and detection accuracy. May Mobility expects that these improvements will benefit communities and riders alike in urban, suburban, and even rural areas. Its autonomy stack includes a third-generation vision stack and next-generation pedestrian modeling that, among other improvements, allow the vehicle to navigate efficiently across many scenarios and interactions.

The newly improved pedestrian modeling feature enables May Mobility’s autonomous system to make more robust predictions about pedestrian behaviors in complex environments. This leads the vehicle to operate more efficiently in crowded environments and respond to pedestrians who are jaywalking or violating the rules of the road. In such environments, new movement policies added to the company’s Multi-Policy Decision Making (MPDM) technology allow the vehicle to adjust its trajectory within the lane to avoid minor impediments.

Also rolled out in the third-generation system are enhanced tele-assist capabilities, which help the vehicle navigate unexpected driving encounters (such as pop-up construction sites). May Mobility’s approach to navigating such encounters pairs its MPDM technology with a remote human operator. While the vehicle moves autonomously throughout this process, the human operator monitors its surrounding environment – ensuring it remains within its operating parameters while approving actions proposed by the vehicle itself. The tele-assist operator can also provide hints to the autonomy system that are used to generate alternative routes through a complex situation.

An upgraded version of the company’s on-vehicle traffic light detection offers increased precision in challenging lighting conditions, such as early mornings and late evenings. The feature also detects non-working traffic lights, leading the vehicle to treat them as stop points.

The partnership has, so far, involved the launch of a highway Navigate on Autopilot (NOA) solution from Pony.ai based on a single Journey 5 computing platform. A prototype vehicle leveraging the both the platform and Pony.ai’s autonomous driving data (captured across dense and complex urban roads) completed highway NOA road tests in China during 2022, and is expected to be production-ready by Q2 2023. Another solution uses a dual Journey 5 platform to enable a NOA experience in cities and similar urban environments, which Pony.ai plans to launch before the end of the year.

The joint solution itself encompasses six cameras, one millimeter wave radar and the Journey 5 processor – offering a unified solution that helps OEMs minimize spending while reducing reliance on HD maps. Based on Pony.ai’s history of complex scene processing, and by considering the interaction between vehicles and their surroundings, the highway NOA function can intelligently navigate in congested driving situations.

In utilizing Horizon Robotics’ technology, Pony.ai aims to combine upstream and downstream value chain partners – forming a smart driving ecosystem with deep integration, efficient collaboration, and mass production. OEMs adopting the solution would then benefit from flexible and convenient smart driving options.

It enables object detection up to 300 meters in real-time under a variety of road and environment conditions, while processing streams from high resolution, high frame rate, cameras.

While Scorpio is currently being evaluated by several OEMs and suppliers, Recogni has said that it can achieve 1000 TOPS at less than 10ms of processing delay with a power consumption below 25 watts. The chip’s efficient design helps it support the flexible design of autonomous driving vehicle stacks while minimizing the impact on driving range. Its short processing time similarly helps the ECU make necessary driving decisions more quickly.

One type of lens will be developed for in-vehicle driver monitoring systems (DMS), while another will be designed for broader ADAS applications. The division has said that both are 20% to 30% thinner than glass-only alternatives, with its design approach enabling it to perform consistently, regardless of temperature, and accurately recognize objects. With its hybrid lenses, LG Innotek is also aiming to broaden the scope of sensor design, with its solution allowing more sensors to be integrated into other body parts, such as the front fillers.

Mass production on LG Innotek’s new camera modules with DMS lenses, and launch camera modules with ADAS lenses, is expected to commence in 2023. The DMS communicates with a camera mounted on the dashboard.

Pilot.Auto is an autonomous driving software platform based on “Autoware”, an open-source software (OSS) for autonomous driving whose development will be led by TIER IV. The platform seeks to help partners achieve their own autonomous driving goals in accordance with TIER IV’s proprietary reference design specifications. Web.Auto is a development and operation (DevOps) software platform for expanding and improving Pilot.Auto to meet individual requirements specified by the partner’s systems. Together, both platforms help benefit partners by reducing development costs, and shortening the time-to-market, of their autonomous solutions.

Pilot.Auto provides a variety of functions for the implementation of autonomous driving systems by customers in their respective communities. Its customization capabilities enable the addition, and expansion, of functions to meet the function and performance requirements for individual use cases and ODD. It can perform evaluations of new functions so that customers can respond to specific requirements within a short time. Pilot.Auto also incorporates Autoware-based autonomous driving systems into a wide range of autonomous vehicles by utilizing TIER IV’s proprietary tools.

The platform allows for the adaptation of software to target driving environments. Here, it offers tuning and verification to ensure safe and comfortable driving in accordance with specific use cases and ODD. It also provides maintenance for customized, integrated, and adapted software, regularly inspecting entire systems – such as vehicles, software, and over-the-air (OTA) software updates.

Pilot.Auto currently provides five variants of the reference designs – cargo transport, shuttle buses, robo-taxis, consumer vehicles, and delivery robots. Each reference design defines the functions and performance requirements that are essential for specific use cases and ODD. The platform is equipped with the basic software and hardware designs to meet those requirements. The reference designs provided by TIER IV provide a foundation for the development of autonomous driving systems for its partners and can significantly reduce the cost and lead time required for the development and operation of autonomous vehicles.

Web.Auto is a cloud-native, edge-intensive, DevOps software platform that aids the development and safe operations of autonomous driving systems. This platform is configured with the necessary development platform tools (e.g., autonomous driving simulators, CI/CD data pipeline) to develop autonomous driving systems efficiently and effectively. It also provides operation platform tools (e.g., fleet management systems, and remote monitoring systems that can be integrated into third-party applications) that support the safe operation and maintenance of vehicles equipped with autonomous driving systems.

TIER IV has formulated a roadmap to develop Pilot.Auto and Web.Auto for the application of the five reference designs. By the end of 2023, it plans to increase the functionality of the cargo transport use design to L4 – operating in closed spaces such as factories, airports, and theme parks. By the end of 2025, the company plans to implement its shuttle bus reference designs for L4 AVs in local cities – implementing the buses across 40 locations in Japan. By this time, TIER IV is also looking to take its cargo transport reference designs overseas while providing L3 autonomy for the consumer vehicle reference design.

By the end of 2028, TIER IV plans to realize the functions of its reference designs for L4 autonomous vehicles – including the shuttle bus, robotaxi, and consumer vehicle – in major urban areas. It will also look to accelerate the application of its reference designs for these vehicles internationally. At the end of the decade, TIER IV plans to be compatible with major driving environments in Japan and overseas for each reference design.

Through Pilot.Auto and Web.Auto, TIER IV ultimately aims to improve the capabilities of autonomous driving development. It will continue to provide autonomous driving software platforms to the market, and form an ecosystem where any organization or individual support the development of autonomous driving technologies and services.

In combining the AI chipmaker’s Hailo-8 processors with Renesas’ R-Car V3H & R-Car SoCs, both companies will power scalable, efficient, and cost-effective ADAS designed for zonal and centralized ECUs. This scalability supports a wide range of ADAS functions ranging from L2+ to L4 autonomous driving functions that require more processing power.

The solution’s own processing power will range from 60 TOPS (tera-operations per second) to 170 TOPS, reserved for multi-sensor perception and fusion operations. Its power efficiency enables vehicle ECUs to be cooled passively, reducing the BOM and system cost, while an open software ecosystem provides OEMs and Tier-1s with more control. The solution works similarly to achieve functional safety for the ADAS/automated driving ECU through Hailo-8’s ASIL B verification.

Within the system, a single Hailo-8 delivers up to 26 TOPS at a typical power consumption of 2.5 W and enables edge devices to run sophisticated Deep Learning applications that could previously only run on the cloud. The processor’s structure further enables higher performance, lower power, and minimal latency – enhancing privacy and reliability for smart devices operating at the edge.

The joint Hailo-Renesas solution is available for customer evaluations and demonstrations now.

Since January 2022, Pony.ai has been conducting road tests optimizing the software to run at low latency and with real-time performance. Pony.ai’s autonomous fleets will feature the NVIDIA DRIVE Orin system-on-chip (SoC) as the centralized compute and AI engine.

Pony.ai’s latest ADC offers single or dual NVIDIA DRIVE Orin configurations, delivering 254 and 508 trillion operations per second (TOPS), respectively. As the world’s highest performance automotive-grade SoC, DRIVE Orin is designed to handle large numbers of applications and deep neural networks simultaneously, while achieving systematic safety standards, to ensure safe, highly advanced intelligent driving.

The ADC, along with Pony.ai’s autonomous driving software and hardware, enables vehicles to intelligently and comfortably maneuver on busy city streets, highways, and expressways. This includes following traffic and merging during congestion; automatic lane-changing to follow navigation; and passing cars. It can also safely handle U-turns; identify traffic signals; and unprotected intersections; nudge around construction zones and accidents; identify and interact with vulnerable road users – as well as drive on mixed-use roads, handle on- and off-ramp merges, and automatically pass electronic toll collection gates, etc.

Pony.ai is one of the first in the AV industry to create a product portfolio featuring multiple configurations with one or more NVIDIA DRIVE Orin processors and NVIDIA Ampere architecture GPUs. This product range enables scalable deployment across self-driving trucks and robotaxis and accelerates Pony.ai’s vision of a robust, mass-production platform for AVs.

Pony.ai develops some of its core hardware in-house. This includes the heterogeneous computing unit based on the NVIDIA chip, which was applied to PonyAlphaX, the company’s fifth-generation autonomous driving system. As of May 2022, Pony.ai has accumulated 13 million (8 million miles) real-world autonomous kilometers, including more than 200,000 driverless kilometers (124,000 miles).

In January 2022, Pony.ai debuted its sixth-generation autonomous driving system design, which has leading-edge sensors, NVIDIA DRIVE compute architecture, and styling and design features for L4 automotive-grade mass production fleets. The first model equipped with Pony.ai’s latest platform, the Toyota Sienna Autono-Maas (S-AM), a seven-seat hybrid electric platform for autonomous mobility, will start road tests in China this year and be deployed within Pony.ai’s public-facing robotaxi operations in the first half of 2023.

Pony.ai began developing DRIVE Orin-based systems with NVIDIA in May 2021, leveraging NVIDIA’s high-performance SoC to accelerate the pace of Pony.ai’s hardware development.

Innovation, commercial application, and a focus on large-scale production form the basis of this partnership. In supporting these goals, RoboSense has already connected with a number of upstream and downstream partners in the industry to integrate advantageous resources and establish a complete supply chain. Deployment of a number of smart production lines of RoboSense in Guangzhou, Shenzhen and other cities are near completion, with an expected annual production capacity of one million units.

Based on this strategic cooperation, the two parties will continue to deepen industrial integration through technological innovation to provide consumers with a safer, more comfortable, convenient and normalized autonomous driving travel experience.

For the pilot service, the Group has obtained a temporary autonomous driving operation permit from South Korea’s Ministry of Land, Infrastructure and Transport. The OEM will collaborate with Jin Mobility, a Korean start-up operating i.M, an AI-powered car-hailing mobility platform. The start-up will be in charge of operating the two IONIQ 5 RoboRide units on its i.M application. Hyundai also plans to expand the pilot service, while further developing autonomous driving technology with consideration for various conditions, such as driving stability. 

Through this pilot, the Group is looking to collect valuable autonomous driving data. Likewise, it plans to further develop the L4 technology trialed within it to navigate safely and flexibly in complicated urban environments. To prepare for such a complicated driving environment, the Group worked with Seoul Metropolitan Government to establish a system that can connect traffic signals with its autonomous vehicles. Since 2018, it has also gathered copious driving data by testing autonomous driving in the Gangnam area.

The Group will also leverage an in-house developed remote vehicle assist system to ensure safety. The system monitors autonomous driving status, vehicle and route, supporting the trip with remote assist functions, such as changing the lane under circumstances where autonomous driving is not feasible. Based on the autonomous technology used in the pilot, a RoboRide vehicle will perceive, make decisions, and control its own driving status, while its safety driver will only intervene under limited conditions.

The RoboRide pilot service will operate from 10 a.m. to 4 p.m., Monday to Friday, to minimize any possible inconveniences on the road. Up to three passengers can partake in a ride, whereas a safety driver will always be present in the vehicle to respond to any emergencies. The Group plans to operate an initial demonstration service for internally selected personnel, and then expand the pilot service targets to general customers in the future.

To commemorate the pilot service, Won Hee-ryong, Minister of Land, Infrastructure and Transport of South Korea, and Oh Se-hoon, Mayor of Seoul City, became the first customers to test drive the RoboRide vehicle.

The agreement between the two firms includes a deep product integration and long-term collaboration roadmap. Within this roadmap, they aim to develop the tools and infrastructure to facilitate the successful deployment of autonomous trucks for Uber Freight’s shipper and carrier customers. Likewise, carriers that purchase trucks equipped with the Waymo Driver will be able to opt-in to Uber Freight’s marketplace through applications that enable them to deploy their autonomous assets on Uber Freight’s network.

These apps aim to streamline several parts of the trucking user experience – including onboarding, load booking and execution, trailer transfers, and payment. At the same time, they also allow for enhanced synergy between human drivers and autonomous trucks, where both can collaborate in a hybrid network to move freight. Both Uber Freight and Waymo Via envision a future where autonomous trucks handle the long-haul portion of driving and drivers shift into short-haul jobs that enhance their quality of life. The companies will further explore what the transfer hub model could look like with Powerloop, Uber Freight’s universal trailer pool program, for efficient transfers between AV trucks and human drivers, streamlining operations between the first, middle and final mile.

Within the long-term agreement, Waymo Via intends to reserve billions of miles of its goods-only capacity for the Uber Freight network. For shippers, the scale and depth of this partnership will allow Uber Freight and Transplace to integrate autonomy into shipper networks. These companies will also advise on how best to adjust their supply-chain strategies for a future where autonomously-driven trucks and human-driven trucks operate side by side.

Ultimately, the hybrid freight model envisioned by Uber Freight and Waymo Via aims to unlock capacity for shippers, increase fuel efficiency, provide carriers with the opportunity to scale their businesses, and streamline global supply chains.

A major highlight will be Kyocera’s new “Road Vehicle Cooperation System,” which contributes to safe autonomous driving by connecting ITS*¹ roadside systems installed in traffic infrastructure with various devices and sensors installed in the vehicle. At the exhibit, we will demonstrate a driving simulator so visitors can experience the Cooperative Road-Vehicle System in action.
*This year’s expo is a hybrid event with online and in-person attendance possible.

Main Exhibit Contents (Planned)
Smart V2I*² Roadside Units
Smart V2I roadside units are designed for installation on utility poles and traffic lights at intersections with poor visibility. This type of device collects information from road infrastructure, such as pedestrian and vehicle locations and traffic light information, and wirelessly communicates the details to vehicles and bicycles. In addition, the Smart V2I roadside units can connect sensors, such as FIR^*3 cameras and other devices. Blind spot information detected by sensors and other devices is transmitted to the vehicle using the V2I units to alert the driver.

FIR Camera Roadside Sensing System
The FIR camera system detects the speed and location of objects in the driver’s path. By imaging the far-infrared radiation these objects emit, FIR camera systems can reveal information the human eye is challenged to detect – especially objects obscured by darkness, rain, fog, or distance. In addition, they can identify not only pedestrians and cars but also bicycles, motorcycles, and large vehicles such as trucks and buses. The FIR Camera Roadside Sensing System helps prevent accidents by communicating information about pedestrians and vehicles in blind spots using the smart V2I roadside units.

Autonomous Bus Demonstration
Kyocera also develops sensors and components to support autonomous driving technology, such as stereo cameras, millimeter-wave radar, and peripheral detection cameras. These technologies will be demonstrated in a 3D hologram of an autonomous bus equipped with Kyocera’s sensors and components, as well as a vehicle control ECU jointly developed with Advanced Smart Mobility Co., Ltd.

Driving Visibility Enhancement Technology
In addition to the above, Kyocera will display two new critical image technologies to improve driving visibility by reducing blind spots.
[Driving Visibility Expansion System]
Displays the most suitable view for the driver and shows the presence of pedestrians, including side- and rear-view angles, in a user-friendly way.
[Optical Camouflage Technology^*4]
Reduces blind spots by making dashboards and pillars appear transparent.

* 1 ITS: Intelligent Transport Systems
* 2 V2I: Vehicle-to-Infrastructure
* 3 FIR: Far Infrared
* 4 Original optical camouflage technology developed in collaboration with Professor Masahiko Inami of the University of Tokyo Advanced Science and Technology Research Center.

With these permits issued by the head office of the Beijing High-level Automated Driving Demonstration Area (BJHAD), ten autonomous vehicles without drivers behind the steering wheel will offer rides to passengers in a designated area of 60 square kilometers in Beijing. These licensed cars will join an existing fleet provided by Apollo Go, Baidu’s autonomous ride-hailing service, in the capital city of China. Starting April 28, 2022, users will be able to hail a driverless ride using the Apollo Go mobile app in daytime from 10:00 to 16:00.

Currently, Baidu has the largest autonomous driving fleet in China. In expanding its driverless vehicle services, Baidu has worked to meet the unique technical challenges of Beijing’s complex traffic environment. The company plans to add 30 more such vehicles at a later stage, expanding its fleet to provide more convenient driverless services to the public.

The new permit represents Beijing’s collaborative and safety-first approach to autonomous vehicle regulation, progressing from the manned autonomous driving stage to the driverless stage. It also represents a benchmark regulation for the global autonomous vehicle industry, given the complexity and high density of urban traffic in Beijing.

Baidu’s success in securing this regulatory permission can be attributed to its strong foundation in AI and its leading test-drive mileage. Baidu has a proven track record of over 27 million kilometers (16 million miles) of road testing accumulated in the past 9 years with zero traffic accidents, including mileage recorded by driverless test cars in multiple cities across China as well as in California.

Arrival has been developing its autonomous driving functionality for the Arrival Van, as part of Robopilot, a project designed to improve the market knowledge, functionality, and public perception of autonomous driving systems. The technology then can be adapted for the planned rollout of all Arrival Vehicles including the Arrival Bus and Arrival Car.

Robopilot is part-funded by Innovate UK and the Centre for Connected Autonomous Vehicles (CCAV). Through the project, Arrival has been able to develop and test its own Automated Driving System (ADS), which uses in-house algorithms, combined with state-of-the-art hardware helping to fast-track the development of self-driving technology and showcase its capabilities. The company is developing a scalable commercial Automated Driving System (ADS), that relies on computer vision and avoids expensive sensing technologies. Arrival is also making HD maps or high-precision GPS completely unnecessary for driving on public roads. Maps will be able to be used as an optional source of data when navigating in a fixed and controlled environment, such as a parcel depot.

Following the successful completion of testing and validation of Arrival’s autonomous driving system in a fixed controlled environment, Arrival will begin testing its autonomous driving functionalities on roads in the UK. Arrival’s aim is for the application to be used in Arrival Vans, across the globe.

New capabilities include:

• Trailering – Customers will be able to trailer their boat, camper and more while driving hands-free.²
• Automatic lane change – The technology can indicate when a lane change is optimal and initiate the maneuver, while following signaling protocols.³
• Enhanced navigation display – Super Cruise compatible roads will be displayed in the in-vehicle navigation systems for vehicles equipped with the embedded Google Maps app and show routes available for hands-free driving, during route selection.

These enhancements will be available on select model year 2022 Super Cruise enabled vehicles, including Cadillac Escalade, Cadillac CT4, Cadillac CT5, Chevrolet Silverado, GMC HUMMER EV Pickup and GMC Sierra.

GM’s vehicle intelligence platform, which provides more bandwidth and data processing power, enabled these enhancements and underscores GM’s software expertise. The digital platform also enables GM to provide over-the-air updates to select model year 2021 vehicle owners, allowing them to take full advantage of Super Cruise’s latest enhancements.

Super Cruise will be available on 22 vehicles by 2023, including Cadillac LYRIQ and GMC HUMMER EV SUV. Additional vehicles that offer versions of Super Cruise include the 2022 Cadillac XT6 and the 2022 Chevrolet Bolt EUV.

GM launched Super Cruise in 2017 with the Cadillac brand and has added capabilities, expanded Super Cruise-enabled routes and availability on more products since then. To date, approximately 10 million miles have been driven by customers using Super Cruise across the U.S. and Canada.

Recently Mobileye applied for and received a New York AV testing permit to drive AVs on New York City streets. Mobileye is the only company currently holding an AV testing permit. During both day and nighttime driving, Mobileye’s AV is getting an advanced AV testing experience. Seven things stand out in particular:

Pedestrians: Jaywalking is common in many cities, but in New York City it is particularly rampant, and is coupled with a high number of pedestrians. An AV must make assumptions about the behavior of those pedestrians and factor those assumptions into its driving policy. Humans do this instinctively; machines must be programmed for it.

Driving behavior: When streets are clogged, drivers become impatient and aggressive. New York City drivers – especially cabbies and other professionals – are much more assertive than drivers in other cities.

Traffic density and road user diversity: Although car ownership in New York City is low compared with other large U.S. cities, the number and variety of road users is especially dense. New York City has more than its share of cabs and limousines, buses, trucks, food carts, horse-drawn carriages, emergency vehicles, bicycles, scooters, skateboards – you name it. 

Double-parking: Who’s parked and who’s not? The question is easier for a human to answer than a machine, and New York City’s population density contributes to a high number of delivery vehicles that must stop to unload. As a result, double-parking is ubiquitous. AVs struggle with this, although the video shows Mobileye’s AVs taking clues from other road users to decide when to maneuver around.

Construction: New York City is one big construction zone, and Mobileye knows this thanks to all the data saved in its always-updating AV maps. While competitors rely on either their own test cars to build maps or spend millions of dollars driving special mapping vehicles, Mobileye receives data about blocked or closed lanes from cars already on the road (data it can and does license back to municipal services, too).

Tunnels and bridges: The island of Manhattan is connected to the surrounding areas via 15 tunnels and 21 bridges, many of which contain narrow lanes framed with bollards or cones – the Achilles heel of many an AV. In the face of all that traffic “furniture” and even multilevel roads, Mobileye’s crowd-source mapping technology and its well-trained sensing system understands all of this and handles it beautifully.

This city really never sleeps (the lights!): Though Paris gets the “city of lights” moniker, Manhattan is electrified at night. The visual noise and light pollution is daunting to an AV’s sensing system. Mobileye AVs handle it easily with only a bit of algorithm tuning.

New York City Event

During a media event on Tuesday, July 19, at Nasdaq in New York City, Mobileye CEO Prof. Amnon Shashua explained how Mobileye technology is preparing for commercial deployment. To make his points, he shared videos of the AV handling New York City’s challenging streets during both day and nighttime driving, as well as during heavy rainfall.