Maps Become Critical Infrastructure for SDVs and Autonomy

 Deon Newman  
SVP and GM of Asia Pacific
HERE Technologies



As the era of software-defined vehicles (SDVs) gathers momentum, the role of digital maps is expanding beyond simple navigation support to become critical driving infrastructure. As ADAS and navigation-on-autopilot (NOA) functions grow more sophisticated, the importance of lane-level maps and real-time location intelligence is also increasing. In an interview with Automotive Electronics Magazine (AEM), Dion Newman, Senior Vice President and General Manager for Asia Pacific (APAC) at HERE Technologies, emphasized that the company supports vehicles in continuously accessing the latest road information through its AI-powered Live Map and integrated software stack, which he described as a key foundation for addressing global regulatory requirements and enabling next-generation SDV architectures.

by BJ Yoon _ bjyun@autoelectronics.co.kr
한글로 보기









When HERE’s map attributes - such as speed limits or lane information - are refreshed, how quickly can these updates be made available for in vehicle functions from a map delivery standpoint?
Newman        At HERE, our priority is to ensure that vehicles operate on fresh, trustworthy, automotive grade map data. Every day, we ingest billions of points of sensor data, drawn from tens of millions of devices and connected vehicles globally, and our AI powered mapmaking pipeline transforms these signals into validated map updates at scale. 

In the SDV era, navigation, ADAS and NOA features depend on much faster and more frequent map updates than traditional systems. As the industry moves from hybrid navigation to fully streaming, software‑defined navigation, automakers increasingly require map content that is refreshed continuously and delivered directly into the vehicle through APIs and integrated software stacks. HERE supports this shift by generating and delivering new map content on a weekly basis through its streaming map APIs, enabling OEMs to bring fresher road, speed‑limit and lane‑level information into their vehicles far more quickly than conventional offline update cycles.


How does HERE ensure the consistency and reliability of its lane level map data when used alongside vehicle sensors in modern perception stacks?
Newman        Ensuring consistent and reliable lane‑level information is one of the biggest challenges in modern mapping, especially as assisted‑driving systems increasingly depend on precise lane geometry and attributes. Over the past two decades, HERE has built advanced AI‑driven mapmaking capabilities that allow us to move from traditional, manual map creation to a highly automated process powered by vehicle‑sensor perception and multi‑source data fusion.

To keep lane‑level content accurate, HERE continuously detects real‑world changes using a wide range of sources including high‑precision lidar from HERE True vehicles, construction‑activity monitoring, vehicle perception data, street‑level imagery and arterial satellite imagery. These signals are processed through AI tools that automatically aggregate, validate and update both physical lane geometry and logical lane attributes, while automated checks and targeted human moderation ensure the final output is consistent, reliable and aligned with real‑world conditions.

This combination of automated perception‑driven updates and multi‑source cross‑validation is what enables HERE to maintain its leadership in global lane‑level map production. It also ensures that leading OEMs can rely on a continuously refreshed, trustworthy foundation for ADAS, NOA and other safety‑critical functions, ultimately giving drivers a more confident, predictable and safer assisted‑driving experience.

 

How does HERE determine which real-world road changes—as seen in rapidly evolving environments like Korea—qualify as significant map change events that should be incorporated?
Newman        We categorize “map‑change events” based on their impact on safety, routing, or driver assistance logic. Korea’s road network evolves quickly due to expressway extensions, continuous construction, lane re‑markings and regulatory changes - which makes prioritization essential.
In markets like Korea, where HERE works in partnership with Hyundai AutoEver (HAE) for map coverage, we focus on identifying the safety‑critical and driver‑assistance‑relevant changes that must be incorporated into the map to keep ADAS and NOA features performing reliably.

HERE uses AI models trained on probe trajectories, sensor observations, HD imagery and historical behavior patterns to flag anomalies such as persistent lane‑level drift signals and unexpected vehicle decelerations. These signals are automatically escalated into change‑event candidates and validated through HERE’s production pipeline.

Additionally, Korea’s dense, multi‑level road structures and frequent temporary traffic interventions (e.g., overnight lane shifts) mean we apply market‑specific thresholds—ensuring that construction‑driven changes, lane boundary shifts, and signage updates are captured and surfaced faster compared to more static markets.


What regional trends are driving demand for real-time, high precision location data in modern vehicles?
Newman        Across APAC, demand for real‑time, high‑precision location data is rising across two‑wheelers, passenger cars and commercial vehicles, driven by stronger safety regulations, fast‑growing telematics adoption, and the expansion of ADAS and emerging automation. 

Governments across the region including Korea, China, India, Japan and Southeast Asia are mandating or encouraging GPS‑based safety, eCall, driver-monitoring and fleet‑monitoring systems.

At the same time, telematics adoption is growing at double-digit rates, with insurers, fleet operators and mobility platforms relying on fine‑grained, time‑stamped location traces to support driver‑behavior analytics, usage‑based services, theft prevention and crash detection. 

This trend is especially pronounced in two‑wheelers and commercial fleets, where food‑delivery and gig‑rider platforms depend on continuous, accurate positioning to optimize dispatching, ETAs and safety interventions, and where shared e‑mobility services require real‑time location for locking, pricing and geofencing.

In passenger vehicles, connected services - including dynamic routing, real‑time traffic, EV‑range guidance, emergency response and contextual in‑car experiences - all rely on always‑on, accurate location information. Collectively, these forces are pushing the entire APAC mobility ecosystem toward more precise, high‑frequency location data as a core enabler of safety, efficiency and next‑generation connected‑vehicle experiences.


How does HERE see navigation evolving as OEMs transition toward software-defined vehicles and connected car ecosystems?
Newman        As the industry transitions toward Software‑Defined Vehicles, navigation is evolving from a standalone feature into core system infrastructure that supports digital cockpits, ADAS, NOA and future automated driving capabilities. 

HERE’s approach centers on delivering a unified live map that ensures all these systems operate from the same, always‑fresh, AI‑powered source of truth. 

This shift is reflected in HERE’s newly announced AI‑powered SDV portfolio, which integrates navigation, driver assistance and autonomy around a single map foundation and helps automakers reduce complexity, accelerate development and deliver cohesive, branded in‑vehicle experiences globally.


What technological innovations have helped accelerate HERE’s map update cycles across fast changing markets like Korea?
Newman        For more than a decade, HERE has deployed commercial applications and amassed a significant patent portfolio for AI and machine learning (ML) techniques used at the core of its digital cartography techniques and location-based services. 

• For example: HERE AI models have the task of extracting map features such as 2D and 3D positioning of road signs. Other models are responsible for validating speed limits in the map. Or trained in detecting and estimating traffic incidents and formulating ETA predictions.  

HERE’s unified mapping architecture and AI-powered map-making capabilities rapidly fuse large volumes of data from thousands of sources – including more than 44 million connected vehicles, sensors and community sources – to create a highly accurate, navigable digital map. This foundation not only supports advanced driver assistance and automated driving systems, but also enables emerging AI-driven cockpit experiences, such as natural language interactions and customized travel planning.

This is reinforced by HERE’s automated feedback loop, which creates a self‑healing map ecosystem: vehicle sensors detect gaps or changes, those signals are aggregated across fleets, and HERE’s automation rapidly generates near real‑time updates - dramatically reducing the need for manual edits and enabling safer, more predictable vehicle behavior on the road. 

On the development side, tools such as SceneXtract accelerate ADAS/AD validation by generating simulation‑ready environments in minutes, helping automakers shorten release cycles while ensuring map freshness at scale.


What role do high-definition and lane level maps play globally in enabling advanced assistance features?
Newman        Around the world, high‑definition and lane‑level maps play a foundational role in enabling safer, more trusted and more intelligent ADAS and NOA features. As assisted‑driving systems mature, they require a unified view of road information by combining sensors, software and location content – allowing the vehicle to understand its environment and make consistent decisions across all driving scenarios.

First, lane‑level maps act as a stable, high‑precision positioning reference, helping ADAS features maintain accurate lane placement, support adaptive cruise control and execute assisted or automated lane changes. They are especially important at decision‑making points such as merges, exits, forks and complex intersections, and they help define the Operational Design Domain (ODD) - clarifying where certain SAE‑level functions can be safely activated.

Second, these maps function as an extended sensor, offering visibility beyond what onboard cameras, radar or lidar can detect. This helps the vehicle anticipate lane merges, geometry changes, connectivity between lanes and typical speeds driven on that stretch of road - supporting smoother, more human‑like driving even in challenging or occluded conditions.

Third, lane‑level maps help ADAS systems understand implicit driving rules and policies. They provide the contextual knowledge a local, experienced driver would have, while supporting compliance with global safety frameworks such as UN Regulation 157 for automated lane‑keeping.

Fourth, high‑definition lane‑level maps offer the most cost‑effective form of redundancy in an ADAS stack. They significantly raise system reliability—augmenting sensor‑based perception—from “good enough” levels to the far higher consistency required for safety‑critical functions.

Finally, these maps serve as the ground‑truth representation of the real world for training end‑to‑end ADAS algorithms and large AI models. They give machine‑learning systems the structured context they need to generalize safely across environments.

This is why nearly every leading OEM and system‑on‑chip provider — across Europe, China, Japan and South Korea — relies on precise lane‑level attributes in their ADAS and NOA deployments. Regardless of their specific technical strategies or algorithms, the industry converges on one point: consistent, lane‑level road intelligence is essential for delivering safer and more predictable assisted‑driving experiences globally.


How does HERE support OEMs in the region with localization, regulatory alignment, and high-frequency updates?
Newman        HERE applies a global standardization plus local customization approach, supported by configurable solutions across data, software, online services, and professional services. Regulatory alignment is emphasized through examples such as EU ISA compliance using HERE ISA Map, and broader compliance focus.

• From the early days of GDPR’s introduction, HERE conducted in-depth research into the regulation and engaged closely with customers, partners, and government agencies to ensure full compliance. As a result, HERE’s navigation maps, connected vehicle products, and technical solutions fully meet GDPR requirements.
• Since July 2024, ISA systems have become a mandatory requirement for all new vehicles entering the EU market — with compliance testing heavily reliant on the accuracy of speed limit data and related services. Thanks to years of technical expertise and deep collaboration with automakers and government bodies, HERE is able to help OEMs achieve scores above 90 in ISA certification tests (well above the passing threshold), providing critical support for compliant product launches in Europe.

Flexible deployment options, including on-premises solutions, are highlighted to address stringent data requirements in certain markets (e.g., local server storage obligations) alongside privacy-by-design principles. 

Real-time updates are supported by large-scale probe and sensor inputs and AI/ML processing designed to keep map content fresh and accurate.

Throughout our journey, HERE has maintained close collaboration with our customers, delivering globally consistent product quality alongside tailored partnership models. Whether it’s technical customization or business model adaptation, HERE provides differentiated capabilities to each automaker and their upstream and downstream partners — including internet companies, manufacturers, technology firms, service providers, and consumer electronics brands. This ability to deliver localized innovation at global scale is central to HERE’s value in helping companies strengthen their product competitiveness and succeed.

For example, in South Korea, geographic and navigation data must be stored on local servers. To meet this requirement, HERE proactively provides planning guidance and tailored solutions to ensure that our partners’ products are fully compliant upon launch. Across the globe, HERE collaborates deeply with automakers and ecosystem partners to ensure their products and services meet local regulatory requirements and can be successfully deployed in diverse markets.


What unique challenges or opportunities does HERE see in delivering high precision location data across APAC markets?
Newman        Across APAC, each mobility segment - passenger vehicles, two‑wheelers and commercial fleets - faces its own set of challenges, but all share a common requirement: real‑time, up‑to‑date, complete (coverage‑rich), and highly accurate location intelligence.

For passenger vehicles: rapidly growing adoption of connected services, ADAS and NOA means automakers need accurate, continuously updated road and traffic information to keep features safe, predictable and compliant across diverse driving conditions.

For the region’s vast two‑wheeler ecosystem: precise location data supports critical use cases such as anti‑theft, rider‑behavior monitoring, food‑delivery and gig‑economy operations, shared e‑mobility, and real‑time geofencing. All of which depend on high‑frequency, sub‑meter‑level positioning in dense urban environments.

For commercial vehicles and logistics operators, the priority is regulatory compliance, dynamic route planning, reliable ETAs and multi‑stop optimization. These depend on continuous location streams that reflect real‑world road changes, restrictions, closures and traffic disruptions.

Because APAC mobility spans everything from scooters weaving through crowded streets to long‑haul trucks, the region demands location data that is fresh, precise and consistent across all vehicle types. This complexity creates a significant opportunity for companies that can provide a unified layer of location intelligence that scales across segments, enabling safer travel, more efficient fleet operations, and better end‑user experiences in one of the world’s most dynamic mobility markets.

To meet these expectations, HERE Technologies has invested heavily in advanced map‑making capabilities that allow us to collect, process, and publish data at scale. This includes: 

• Vehicle perception data (VPD) from connected vehicles to update road geometries and road attributes. 
• Address and POI updates sourced through feeds from ride‑sharing platforms and large‑scale delivery companies. 
• Local authoritative sources and regional partners, which further contribute essential datasets to enhance map completeness, accuracy, and compliance across markets. 

Together, these inputs allow HERE to continuously and reliably update both static and dynamic map content across APAC’s diverse and rapidly evolving mobility landscape.


What major innovations are HERE prioritizing in the coming years for SDV platforms, real-time navigation, or sensor-enhanced mapping?
Newman        Over the next three to five years, HERE is focused on advancing an AI‑powered live map and a unified SDV‑ready software stack that help automakers accelerate development, simplify integration, and deploy safer, more intelligent driving experiences globally. 

HERE is also investing in next‑generation tooling such as SceneXtract - an AI‑driven simulation engine that reduces scene‑generation time from weeks to minutes and expanding capabilities within the HERE SDV portfolio, including enhanced lane‑level rendering, ADAS GUIs, and Behavioral Maneuvers for smoother, more human‑like automated driving. 

Additionally, HERE’s Automated Driving Zones (ADZ) provide OEMs precise control over where and how automated features activate, supporting safe ODD management at scale. 

Collectively, these innovations are designed to help automakers respond to rising expectations for automation, deliver globally compliant features, and build future‑ready SDV architectures with far greater speed and efficiency.


In the era of fully software-defined self-driving vehicles, how do you see the role of high-definition maps evolving?
Newman        As vehicles become increasingly software‑defined and progress toward higher levels of automation, the role of maps is expanding from navigation support to critical driving infrastructure. Even as perception systems advance, onboard sensors alone cannot provide the foresight, redundancy or contextual awareness required for safe, predictable automation. HERE’s AI‑powered live map delivers this missing layer by providing detailed lane topology, road geometry, curvature, speed limits and other essential attributes that allow the vehicle to anticipate the road ahead and operate more smoothly and safely. 

In practice, this means HD and lane‑level maps are becoming more - not less - important. They reduce compute load by giving vehicles a prior understanding of road context, strengthen overall safety by complementing sensors in adverse weather or complex road environments, and give OEMs the flexibility to define and control Operational Design Domains (ODDs) consistently across markets through tools like HERE Automated Driving Zones (ADZ).

At the same time, HERE’s unified live‑map architecture ensures that cockpit navigation, ADAS, NOA and automated‑driving features all consume the same, always‑fresh data, which is essential for reliable behavior at scale across global deployments. 

Ultimately, as OEMs move into the SDV era, maps evolve into a core enabler of intelligent mobility, providing the real‑time, high‑precision location intelligence that automated systems rely on to make safer, more human‑like decisions worldwide.


How does HERE leverage AI?
Newman        For drivers, the next wave of innovation is about vehicles that feel more intuitive, more helpful and more proactive - not just more automated. Agentic AI makes this possible by giving in‑car systems the ability to understand context, anticipate needs and take action on behalf of the user. Whether it’s helping a driver plan their commute, suggesting where to charge an EV, or assisting with errands and commerce in the car, the experience becomes smoother and more personal because the system knows where the driver is, where they’re going and what they typically do.

HERE is helping customers design and co‑create agentic AI solutions that use location intelligence as the catalyst for real‑world decision‑making. In mobility and logistics, for example, HERE is building AI agents that automate planning, execution and analysis, enabling real‑time decisions, adaptive operations and seamless collaboration between agents, users and systems. HERE provides these AI agents with the vital location context they need for greater awareness, proactivity and control, while feedback loops ensure solutions stay effective and continuously improve over time.

HERE uses AI/ML powered software to identify millions of map changes and road network events every day, including those specifically relevant for e-commerce logistics.  

HERE delivers AI-powered solutions across automotive and MI, including:  

• SceneXtract: Generates high-fidelity ADAS and AD simulation scenes easily and in minutes using AI. Accelerate your SDV development and unlock new potential for simulation and testing 
• HERE AI assistant: AI-powered voice-driven solution that adapts in real time, enhances driver engagement, and scales affordable across vehicle platforms. 
• SDV Accelerator development: A comprehensive platform incorporating AI services for Software-Defined Vehicle (SDV) 
• Through integrations like Alexa+ with HERE Navigation, drivers can ask conversational questions — “Find me parking near my destination,” or “Stop for coffee along the way” — and get answers that combine voice intelligence with precise, real‑time road context. This makes everyday tasks safer and easier because drivers can keep their hands on the wheel while the system handles planning, searching and route adjustments.

[AEM] Automotive Electronics Magazine

---