Skip to main content
XenonStack

Enterprise AI

XenonStack Blog Automotive IoT App Development for Connected Vehicles

and | Published | Updated

Key takeaway

  • Automotive IoT app development in 2027: the connected vehicle data path, use cases to fund first, R155 and R156 rules, and how to sequence adoption.

Automotive IoT app development is the work of connecting the sensors and electronic control units in a vehicle to apps that drivers, fleet managers and engineers can act on. In 2026 that means three things built together: a vehicle data path (sensors, gateway, edge and cloud), an app surface (a phone app, an in-vehicle app on Android Auto or Android Automotive OS, or a fleet console), and the security and software update controls that regulators now audit.

This guide is for leaders who decide what to build and in what order. It covers the connected vehicle app stack, the use cases worth funding first, what has changed in platforms and regulation, and when not to build. For the driving-automation side of the stack, see our guide to ADAS and autonomous vehicles, and for the maintenance models that sit on top of vehicle data, see AI predictive maintenance for vehicles.

What is automotive IoT app development?

An automotive IoT app reads signals from the vehicle (location, speed, engine and battery health, load, fault codes) and turns them into information or actions for a person or another system. The vehicle side has sensors, control units and a telematics or gateway unit that connects to the network. The app side ranges from a driver's phone app that unlocks doors or pre-heats the cabin, to a fleet platform that tracks hundreds of vehicles and raises maintenance work.

Most programmes involve four kinds of software. Choosing which one you are building, and who will use it, settles most later design decisions.

App typeWhat it doesWhen it fits
Driver companion app (phone)Remote functions such as lock, climate and charge status; trip and service remindersConsumer or employee-driver vehicles where the phone is the main touchpoint
In-vehicle app (Android Auto or Android Automotive OS)Navigation, points of interest such as parking, charging and fuel, media, weather and IoT controls on the car displayWhen the driver needs the information while driving and the vehicle supports the platform
Fleet and telematics consoleLocation tracking, route adherence, idle time, load and vehicle health across a fleetLogistics, field service, public transport and rental fleets
Connected vehicle back endCollects, normalises and stores vehicle data; runs analytics, alerts and software update campaignsEvery programme needs one; the question is whether to build, buy or extend

How does data move from the vehicle to the app?

Vehicle data passes through layers, and each layer is a design decision. Sensors and control units produce signals; a gateway or telematics control unit reads them, often through the standard on-board diagnostics port in older or retrofitted fleets; an edge layer filters and pre-processes; the cloud stores, analyses and serves the app. Sending every signal to the cloud is expensive and slow, so the common pattern is to decide at the edge what is worth sending. Our overview of edge computing for IoT explains that trade-off in more depth.

Two-column diagram: vehicle data layers (sensors, gateway, edge, cloud and apps) on the left, matched to decisions on the right (signals, fitting, edge rules, approval under UN R155 and R156)

The hardest part is usually not connectivity but meaning. Different models, suppliers and control units name the same signal differently. The COVESA Vehicle Signal Specification (VSS) addresses this with an open, hierarchical data model that gives vehicle data one shared meaning in the vehicle and in the cloud, and it is designed to be protocol and technology agnostic. Mapping your signals to a common model early saves rework when a second vehicle make or supplier joins the fleet. The same layered thinking applies to city-scale IoT, as covered in our data processing in IoT architecture post.

LayerWhat happensDecision to make
Sensors and control unitsRaw signals: position, speed, temperature, battery, axle load, fault codesWhich signals the use case actually needs
Gateway or telematics unitReads the vehicle network or diagnostics port and connects to the mobile networkFactory-fitted unit or retrofit device
Edge processingFilters, aggregates and triggers events close to the vehicleWhat to decide locally and what to send
Cloud platformStores and normalises data, runs analytics, alerts and update campaignsBuild, buy or extend; data model; retention
Apps and business systemsDriver app, in-vehicle app, fleet console, maintenance and ERP integrationWho acts on each signal, and how fast

Which use cases should enterprises build first?

The use cases in the original version of this guide are still the core of most programmes. What has changed is that each one now needs a clear owner for the action, not just a dashboard.

Use caseWhat it doesWhen it fits
Fleet location and route trackingUses GPS, accelerometer and diagnostics data to report location, speed changes, idle hours and route adherenceAny fleet where late or off-route vehicles cost money
Vehicle health and maintenance alertsWarns drivers and fleet owners of upcoming service needs and faults before a breakdownFleets with high downtime cost; pairs with predictive maintenance models
Weight and load monitoringAxle-mounted sensors report real-time load on trucks and trailers, including liquid and bulk cargoHaulage, chemicals, fuel and bulk logistics
Traffic and journey informationDelivers traffic, road closure and route information to drivers or driving systemsDelivery, field service and passenger transport
Remote vehicle functionsLock, climate, lights and charging control from a phone appConsumer vehicles, car sharing and rental
Software updates over the airDelivers and records software updates to vehicles in the fieldAny vehicle with updatable software; regulated under UN R156 where type approval applies

Maintenance is often the best first case because the cost of a breakdown is easy to measure and the action is clear: create a work order, book the vehicle in, order the part. The value comes from closing that loop from a signal to an approved action, not from the alert alone. XenonStack teams see the same pattern across automotive operations and supply chain work: the data is usually available, but the decision about what to do with it is spread across people and systems.

Video: how AI agents take an asset maintenance finding through to an approved work order. The example is general asset maintenance; the same loop applies to vehicle health alerts.

What has changed in automotive IoT since this guide was first written?

ChangeWhat it meansCurrent status
Software-defined vehiclesMore vehicle functions are delivered and changed through software, so the app and the vehicle software lifecycle are now linkedIndustry direction; VSS is positioned to support software-defined and connected-vehicle ecosystems
Cyber security becomes type-approval lawManufacturers need a certified cyber security management system covering suppliersUN R155: EU new types from 6 July 2022, all existing type approvals from 7 July 2024 (UK VCA)
Software updates are regulatedUpdates need a software update management system and recordsUN R156: EU new types from July 2022 and July 2024; new complete vehicles from 7 July 2026 (UK VCA)
Built-in car platformsApps can run natively on Android Automotive OS as well as projected through Android AutoSupported by Google's Android for Cars programme, with category rules for driving and parked use
Managed vehicle data services shiftAWS IoT FleetWise closed to new customers on 30 April 2026AWS points new customers to Guidance for Connected Mobility on AWS
Common vehicle data modelsShared signal names across vehicle and cloudCOVESA VSS is an open, widely adopted model

Which in-vehicle app platform fits which need?

Google's Android for Cars documentation separates two platforms. Android Auto projects an app from the driver's phone onto the car display and needs a phone running Android 9 or later. Android Automotive OS is a standalone Android system built into the car, so the app is installed in the vehicle. Google recommends one app architecture that works for both.

Supported categories matter for scope. Navigation, points of interest (parking, charging and fuel), IoT control and weather apps are built with the Android for Cars App Library and run on both platforms. Video and browsers are limited to Android Automotive OS and only while parked, and games are parked-only on both. If your use case needs the driver to see rich content while moving, check the category rules before you design the screens.

PlatformWhere the app runsWhen it fits
Android AutoOn the phone, projected to the car displayReaching drivers across many existing vehicle makes without OEM integration
Android Automotive OSInstalled in the vehicle's built-in systemOEM or fleet programmes where vehicles ship with the platform and no phone should be required
Fleet console (web and mobile)Outside the vehicle, for operatorsMost enterprise fleet use cases; no in-car display needed

What governance, security and cost issues apply?

Security is now a compliance requirement, not only good practice. The UK Vehicle Certification Agency explains that UN R155 and R156 are audit-based: manufacturers submit their cyber security and software update management systems for assessment, and must cascade security requirements to suppliers, usually through contracts. Any app or back end that can change vehicle behaviour or deliver updates falls inside that scope. ISO/SAE 21434:2021 sets the cybersecurity engineering process for vehicle electrical and electronic systems from concept to decommissioning, including software and communication interfaces.

For enterprise fleets that do not build vehicles, the obligations fall differently but the questions are similar: who can send a command to a vehicle, how is it approved, and can you show afterwards what was decided and why. ElixirData Context OS keeps governed context, versioned policy and decision traces in one decision layer above AI agents, so actions triggered from vehicle data follow policy. The Audit Agent assembles decision lineage, policies and approvals for compliance traceability.

Cost follows data volume. Collecting every signal at high frequency drives connectivity, storage and processing costs without improving decisions. AWS describes the intended pattern for its vehicle data service as sending only the data needed for analysis, which is a sound rule for any platform. Decide signals and sampling rates per use case, process at the edge where latency matters, and review retention. NexaStack is an agentic operating platform for physical AI that runs across on-premises, cloud and hybrid environments, which suits teams that need edge decisions with central governance.

How should an enterprise sequence adoption?

  1. Pick one decision. Choose a use case with a measurable cost, such as unplanned downtime, route adherence or load compliance, and name who acts on it.
  2. Agree the signals and data model. List the signals the decision needs and map them to a common model such as VSS before connecting more vehicles.
  3. Choose the data path. Factory-fitted or retrofit telematics, edge processing rules, and a back end you build, buy or extend. Check the current status of any managed service.
  4. Set security and update controls. Define who can send commands and updates, how they are approved and how they are recorded, in line with R155, R156 and ISO/SAE 21434 where they apply.
  5. Build the app surface last. Phone app, in-vehicle app or fleet console, chosen by who acts and where they are.
  6. Prove the loop, then scale. Measure the decision outcome on a pilot fleet before adding use cases or vehicle makes.

What are the trade-offs, and when should you not build a custom app?

Building a custom connected vehicle app gives control over data, user experience and integration, but it also means owning device compatibility, security assessment, update handling and long-term support. That is justified when vehicle data is central to your service or your fleet mix is unusual. It is hard to justify when an existing telematics provider already covers your vehicles and use cases.

SituationBetter choice
Standard fleet tracking on common vehicle makesUse an established telematics service and integrate its data
Vehicle data drives your core service or productBuild the back end and data model; buy commodity parts
In-car display needed while drivingDesign within Android for Cars category rules from the start
You cannot yet show who approves vehicle commandsDo not ship remote control features until approval and records are in place

The next step for most teams is a working session on the first decision rather than a platform choice. Our IoT application development services and real-time analytics services cover the build side.

To map one vehicle-data decision your agents should make, and the context and policy it depends on, book a Context Workshop. See what happens in the workshop.

What should you plan for in 2027?

If you sell or operate vehicles in Great Britain, the UK government's consultation outcome (last updated 16 September 2025) sets planned dates for UN R155 and R156.

  • 1 June 2027: planned date for complete and incomplete vehicles under both R155 and R156.
  • 1 June 2028: planned date for completed vehicles under R155, with R156 following on 7 July 2029.

Check the current position with your type approval authority, then make sure any app or back end that can send commands or updates has approvals and records in place before those dates.

Frequently asked questions

What is automotive IoT app development?

It is the work of building the software that connects vehicle sensors and control units to drivers, fleet operators and business systems. It covers the in-vehicle or phone app, the telematics data path through a gateway or edge layer, and the cloud services that turn vehicle signals into actions such as alerts, work orders or over-the-air updates.

What is the difference between Android Auto and Android Automotive OS?

Android Auto projects an app from the driver's phone onto the car's display, so the app is installed on the phone. Android Automotive OS is built into the vehicle as a standalone Android system, so the app is installed in the car. Google recommends one app architecture that works for both.

Which regulations apply to connected vehicle software?

In markets that apply UNECE type approval, UN R155 requires a certified cyber security management system and UN R156 requires a software update management system. In the EU, R155 applied to all existing vehicle type approvals from July 2024. ISO/SAE 21434:2021 sets the cybersecurity engineering process across the vehicle lifecycle.

Is AWS IoT FleetWise still available for new projects?

AWS states that IoT FleetWise is no longer open to new customers from 30 April 2026 and points to Guidance for Connected Mobility on AWS as the alternative. New programmes should check current service status before committing to any managed vehicle data service.

Why use a common vehicle data model such as VSS?

COVESA's Vehicle Signal Specification gives vehicle data one shared meaning across the vehicle and the cloud through a hierarchical tree of signals. Using it means apps, analytics and suppliers read the same signal names, which cuts the mapping work that fragmented vehicle data otherwise creates.

Where should an enterprise start with automotive IoT apps?

Start with one decision that has a measurable cost, such as unplanned fleet downtime or route adherence. Agree the signals it needs, collect only those, set the security and update controls, and prove the loop from signal to action before adding more use cases.

Conclusion

Automotive IoT apps have moved from novelty features to regulated, data-driven services. The programmes that work pick one decision, agree the vehicle data model, put security and update controls in place, and only then build the app surface. Start small, prove that a vehicle signal leads to a better and recorded action, and scale from there.

Follow XenonStack on LinkedIn: XenonStack on LinkedIn for new guides on agentic AI, data and platform engineering.

Ready to put AI agents to work? Explore ElixirClaw, the agentic OS platform to build, orchestrate and govern AI agents.

Table of Contents

navdeep-singh-gill

Navdeep Singh Gill

Global CEO and Founder of XenonStack

LinkedIn profile

Navdeep Singh Gill is serving as Chief Executive Officer and Product Architect at XenonStack. He holds expertise in building SaaS Platform for Decentralised Big Data management and Governance, AI Marketplace for Operationalising and Scaling. His incredible experience in AI Technologies and Big Data Engineering thrills him to write about different use cases and its approach to solutions.

Get the latest articles in your inbox

Subscribe Now

Interested in solving your challenges with the XenonStack team?
Get Started

Tell us about you and your priorities so we can prepare the right conversation. It takes about 2 minutes.

Please Select your Industry
Banking
Fintech
Payment Providers
Wealth Management
Discrete Manufacturing
Semiconductor
Machinery Manufacturing / Automation
Appliances / Electrical / Electronics
Elevator Manufacturing
Defense & Space Manufacturing
Computers & Electronics / Industrial Machinery
Motor Vehicle Manufacturing
Food and Beverages
Distillery & Wines
Beverages
Shipping
Logistics
Mobility (EV / Public Transport)
Energy & Utilities
Hospitality
Digital Gaming Platforms
SportsTech with AI
Public Safety - Explosives
Public Safety - Firefighting
Public Safety - Surveillance
Public Safety - Others
Media Platforms
City Operations
Airlines & Aviation
Defense Warfare & Drones
Robotics Engineering
Drones Manufacturing
AI Labs for Colleges
AI MSP / Quantum / AGI Institutes
Retail Apparel and Fashion

Continue