Intelligence at source
Process physical signals beside the sensor, where context is fresh and response time matters.
Olive technology platform
From electrons to autonomous motion.
Olive co-designs electronics, embedded compute, real-time software, physical AI, precise timing and ROS 2 autonomy as one deployable technology stack—from a sensing element to a complete AMR.
Robotics electronics
Real-time olixOS™
Native ROS 2 / DDS
Full AMR stack
OLIVE EDGE ARCHITECTURE / LIVE DATA PATHSYNCHRONIZED
olive® EDGESiP / SoMup to 6 TOPS
PHYSICAL / 01electronics + I/O
RUNTIME / 02olixOS™
INTELLIGENCE / 03olixAI™
SYSTEM / 04ROS 2 autonomy
FORM FACTOR38 × 38 mm
AI COMPUTEup to 6 TOPS
EXECUTIONreal-time edge
SYSTEM SCOPEsensor → AMR
Engineered as one system
Hardware and software, co-designed.
Robotic performance is limited by the weakest boundary between power, signal integrity, compute, timing, software and motion.
Olive designs across those boundaries. Power rails, sensor buses, DMA buffers, kernel scheduling, inference, DDS communication and autonomy are treated as one measurable path. The same engineering logic scales from an embedded sensor node to a complete mobile robot.
One temporal truth
Preserve timing from hardware acquisition through fusion and DDS publication.
Native interoperability
Expose devices directly as ROS 2 nodes over standards-based DDS—not through proprietary wrappers.
Platform reuse
Carry one hardware and software architecture from prototype sensors to production robot fleets.
The Olive stack
Six layers. One robot.
Select a layer to trace the system from board-level electronics and embedded execution to physical AI, ROS 2 communication and autonomous behavior.
Layer 01 / System electronics
The physical foundation for deterministic data.
Power integrity, clock quality, signal routing, EMC behavior and interface protection define what the software can trust. Olive treats the PCB and embedded runtime as one timing-aware system.
PowerSequenced multi-rail design
InterfacesSensor + industrial I/O
ValidationSI / PI / EMC aware
Layer 02 / Embedded hardware
A compact computing foundation built beside the sensor.
Olive SiP and SoM platforms integrate compute, memory, power management, flexible I/O and AI acceleration in a 38 × 38 mm footprint—forming a reusable hardware core for intelligent robotic devices.
Footprint38 × 38 mm
AI accelerationUp to 6 TOPS
ArchitectureSiP or SoM integration
Layer 03 / Operating system
Deterministic execution from driver to data stream.
olixOS™ is Olive’s embedded real-time Linux foundation, bringing hardware abstraction, timing services, optimized scheduling and the ROS 2/DDS runtime into a controlled device environment.
KernelReal-time Linux
Data movementDMA / zero-copy HAL
NetworkingPTP / TSN capable
Layer 04 / Physical AI
Models that understand signals, states and change over time.
olixAI™ runs inference on continuous sensor data at the edge—from adaptive filtering and calibration to sensor fusion, anomaly detection and confidence-aware state estimation.
InputTime-series sensor data
FunctionsFusion / anomaly / state
OutputSignal + confidence
Layer 05 / Robot interface
The sensor is the ROS 2 node.
Each Olive device publishes directly through a DDS-backed ROS 2 interface with topics, timestamps, services and configurable QoS. No host-side wrapper is required to make the device robot-ready.
ProtocolDDS / RTPS
SemanticsROS 2 topics + services
IntegrationNo custom wrapper
Layer 06 / Robot autonomy
Close the loop from perception to physical motion.
At robot level, synchronized sensing feeds localization and SLAM, costmaps and planning, holonomic or differential-drive control, safety supervision and fleet operations through one ROS 2 graph.
PerceptionLiDAR / vision / IMU
AutonomySLAM / Nav2 / control
OperationsWeb UI / OTA / fleet
Electronics + embedded systems
Engineered below the software boundary.
Reliable robotics starts before the kernel boots. Olive develops the power tree, clocks, sensor buses, high-speed interfaces, memory, compute and board support package as one measurable embedded platform.
EMBEDDED BOARD / LIVE NET EXPLORERDESIGN VIEW
POWERPMIC + rails
PHYSICAL I/Osensor buses
COMPUTE DOMAINCPU · NPU · DMA
MEMORYDDR + storage
NETWORKGbE · USB · PHY
TIMEBASEclock / PTP
DOMAIN / POWER INTEGRITY
Stable rails before stable inference.
Sequencing, transient response, decoupling and power-domain supervision protect sensor accuracy and compute stability across boot, load steps and network activity.
ENTRYprotected input
CONTROLPMIC sequencing
MEASURErail telemetry
GUARDbrownout state
Route for the waveform.
Controlled impedance, return paths, length constraints and connector strategy preserve high-speed links and low-level sensor signals.
Meet every physical layer.
SPI, I²C, UART, CAN, RS-485, USB and Gigabit Ethernet connect sensors, actuators, service tools and robot networks.
Hardware described in software.
Device tree, pin control, clocks, DMA, IRQ routing and kernel drivers make the electrical design deterministic and inspectable.
Performance in the enclosure.
Power filtering, ESD protection, grounding, shielding and thermal paths are considered with PCB geometry and sustained workloads.
Designed for production.
Test points, programmed identity, boundary checks, calibration records and traceable diagnostics support manufacturing and service.
Separate critical authority.
Watchdogs, hardware enables, power-state monitoring and isolated control paths keep supervisory functions explicit.
Deterministic signal path
From physics to robot state.
Every stage preserves the information robotics depends on: precise time, physical context, model confidence and deterministic delivery.
Acquire
Capture high-rate physical measurements through optimized sensor interfaces and buffered I/O.
Timestamp
Attach acquisition-time context and align channels against a shared device or network clock.
Process
Calibrate, compensate, filter, resample and validate the incoming signal close to the source.
Infer
Run embedded fusion, time-series inference, anomaly detection and confidence estimation.
Publish
Deliver raw and intelligent outputs as native ROS 2 messages with purpose-defined DDS QoS.
AcquisitionBuffered sensor interfaces
TimingHardware-assisted timestamping
ExecutionOn-device AI pipeline
DeliveryDeterministic DDS QoS
Software-defined hardware
Real edge intelligence in 38 × 38 mm.
A reusable SiP/SoM architecture that places capable compute directly inside the sensing product.
olive®EDGE COMPUTE CORE
6 TOPSUp to AI performance
38 × 38 mmCompact platform size
Compute where data begins
Less movement. More meaning.
Sending every raw sample to a central computer adds bandwidth, latency and integration work. Olive moves the first intelligence layer to the sensor edge, so the robot receives synchronized, calibrated and context-aware data.
01
Integrated AI acceleration
Run quantized signal-processing and inference workloads with up to 6 TOPS of on-device compute.
02
Flexible hardware interfaces
Bridge sensors, timing sources and robot networks through one consistent embedded architecture.
03
Local raw + intelligent outputs
Keep access to source measurements while adding fused state, events and model confidence.
04
Platform-level reuse
Apply the same compute core, operating system and ROS-native interface across multiple products.
Time as infrastructure
Every signal on the same clock.
Fusion is only as trustworthy as its timestamps. Olive carries time through the complete path—from acquisition and buffering to inference and DDS publication—while supporting network synchronization through hardware-assisted IEEE 1588 PTP and TSN-ready communication.
MULTI-SENSOR TIME FABRIC / t = 178453.420921PTP LOCKED
CHANNEL 01IMU / 1 kHz
SYNC +0.4 μsCHANNEL 02F/T / 500 Hz
SYNC -0.2 μsCHANNEL 03RADAR / 50 Hz
SYNC +0.7 μsCHANNEL 04ENCODER
SYNC +0.1 μsSHARED TIME DOMAINAcquisition → olixOS™ → olixAI™ → DDS
microsecond-class alignmentBuilt on olixOS™
The real-time core for embedded robotics.
olixOS™ turns compact embedded hardware into a deterministic, maintainable and robot-native device platform.
olixOS™ / SYSTEM STATUSRT ONLINE
olive@edge:~$ olix-status --realtime
[kernel] PREEMPT_RT active
[timing] ptp clock locked
[hal] sensor pipeline zero-copy
[dds] native endpoint ready
olive@edge:~$ _
EXECUTIONCPU affinity + IRQ isolation
DATA PATHDMA-driven sensor HAL
NETWORKPTP + TSN services
MIDDLEWAREROS 2 executor + DDS QoS
Real-time scheduling
Preemptible execution, CPU affinity and interrupt control for predictable sensor workloads.
Zero-copy HAL
Optimized device access and DMA-oriented data paths reduce unnecessary copies and context switches.
Deterministic networking
DDS QoS, PTP timing and TSN-ready networking keep critical streams prioritized.
Controlled device lifecycle
A unified runtime for configuration, diagnostics, updates and product-level reliability.
ROS inside
Native from the first packet.
With Olive, ROS 2 is not a host-side adapter added after the device is built. The device itself participates in the robot graph, publishes through DDS and exposes a consistent interface from development to deployment.
✓Topics, services, parameters and timestamps directly on-device
✓Configurable reliability, durability, history and deadline QoS
✓Discovery and communication through standards-based DDS / RTPS
✓Same interface for one robot, a test bench or an entire fleet
ROS 2 COMPUTATION GRAPHDDS DISCOVERY ACTIVE
EDGE NODE/olive_sensor
MIDDLEWAREDDS / RTPS
CONSUMER/navigation
CONSUMER/controller
CONSUMER/data_logger
$ ros2 topic echo /olive/sensor/data --once
stamp: 178453.420921 frame_id: olive_link
status: synchronized confidence: 0.998
Control-loop-aware execution
Separate the critical path.
Robotic sensing is not one generic workload. Olive structures acquisition, conditioning, inference, publication and device management as distinct execution domains, so time-critical work stays protected from background activity.
OLIXOS™ EXECUTION TRACE / ILLUSTRATIVEDEADLINE MONITORED
DOMAIN / 00IRQ + DMA
HARD RT
DOMAIN / 01Acquisition
PINNED
DOMAIN / 02Signal pipeline
BOUNDED
DOMAIN / 03AI inference
ACCELERATED
DOMAIN / 04ROS 2 publish
QoS AWARE
DATA PLANEAcquisition → state output
MANAGEMENT PLANEConfig → logs → updates
GUARDRAILSAffinity → priority → health
Protect acquisition
Use interrupt control, CPU affinity and DMA-oriented I/O to keep high-rate sampling independent from noncritical services.
Budget every stage
Treat filtering, fusion, inference and publication as measurable stages in one end-to-end timing path.
Expose system health
Make timing state, synchronization, data quality and runtime status visible to the robot—not hidden inside a driver.
Semantic sensor data
A sample is more than a number.
Robotic software needs source time, coordinate context, calibration state, uncertainty and device health alongside the measurement. Olive carries that meaning from acquisition to the ROS graph.
SIGNAL CONTRACT / ILLUSTRATIVE PAYLOADSTREAM ACTIVE
header.stamp178453.420921
sequence00018432
frame_idolive_sensor_link
signal[][ +0.021, −0.994, +0.113, … ]
covariance[][ 2.1e−5, 0, 0, … ]
time_statePTP_LOCKED / SOURCE_TIME
modelolixAI / confidence: 0.998
device_healthNOMINAL
Transport matched to intent.
High-rate state, fault events and device configuration have different delivery needs. Olive can apply purpose-defined DDS behavior instead of forcing every signal through one generic channel.
High-rate stateKEEP_LAST / LOW LATENCY
Fault + eventRELIABLE / ORDERED
ConfigurationRELIABLE / ACKNOWLEDGED
Fleet discoveryDDS / RTPS NATIVE
Designed to be inspected.
Engineering teams can reason about the complete path through explicit status and diagnostics.
SYNC STATEDROPPED SAMPLESMODEL VERSIONDATA QUALITYDEVICE HEALTHQOS STATUS
A different integration model
From peripheral to participant.
Olive changes the sensor’s role—from a passive data source that needs a host-side software stack into an intelligent network node.
SYSTEM LAYER
Conventional sensor
Olive-native device
Compute
Raw measurements sent to a host for processing
Edge processing
Signal conditioning and inference on-device
Signal conditioning and inference on-device
Software
Driver, SDK and wrapper maintained per host platform
olixOS™ runtime
Controlled hardware-to-middleware stack
Controlled hardware-to-middleware stack
Timing
Arrival-time stamps and independent device clocks
Source-time context
Hardware-assisted timestamping and PTP
Hardware-assisted timestamping and PTP
Interface
Vendor protocol translated on the robot computer
Native ROS 2 / DDS
Topics and QoS published by the device
Topics and QoS published by the device
Scale
Integration effort grows with every device type
Reusable platform
One architecture across products and fleets
One architecture across products and fleets
Full robotics stack
From a device node to a complete AMR.
The same architecture expands into olixRobot™ ANT1: synchronized perception, state estimation, SLAM, Nav2 planning, motion control, safety I/O and fleet operations in one ROS 2-native system.
AMR DIGITAL TWIN / MAP FRAMEAUTONOMY ACTIVE
ANT1 / base_link
POSE x:4.82 y:2.17 θ:-0.04GOAL / DOCK_03FLEET / UNIT_07
LOCALIZATIONmap → odom → base
PLANNERglobal + local
CONTROLholonomic x · y · θ
OPERATIONSROS 2 · Web UI · OTA
STACK / MULTIMODAL PERCEPTION
Give the robot a synchronized world.
LiDAR, cameras, IMU and wheel encoders enter the graph with source time and coordinate context. The perception layer validates data quality before it reaches localization or safety logic.
01SenseLiDAR · vision · IMU · encodersSOURCE TIME
02LocalizeEKF · odometry · SLAM · TFSTATE + COV
03Plancostmaps · behavior tree · Nav2PATH + GOAL
04Controlkinematics · trajectory · drivesCMD_VEL
05Operatehealth · fleet · Web UI · OTALIFECYCLE
Technology in physical form
One architecture. Multiple robot nodes.
Explore how the platform appears as inertial intelligence, synchronized industrial sensing, robot connectivity and a complete autonomous mobile robot.
PTP SYNCHRONIZED
OLIXSENSE™ / INDUSTRIAL IMU
X2 Ultra
PoE Ethernet, IEEE 1588 PTP, redundant Multi-MEMS sensing and embedded Physical-AI fusion for time-critical robotics.
FUSION ACTIVE
OLIXSENSE™ / EMBEDDED IMU
X1 Pro
Compact redundant motion sensing with embedded fusion, native ROS 2 and Ethernet-over-USB through one Type-C connection.
NAV2 / AUTONOMY ACTIVE
OLIXROBOT™ / AMR PLATFORM
ANT1
Holonomic mobility, multimodal perception, ROS 2/DDS, Nav2 and edge AI in a modular autonomous robot platform.
5G LINK ONLINE
OLIXLINK™ / ROBOT CONNECTIVITY
C1
Industrial 5G, Wi-Fi 6, dual-SIM resilience, PoE+ Ethernet and ROS 2 diagnostics for connected robots and fleets.
MOTIONX1 Pro / fused IMU state
PRECISION TIMEX2 Ultra / PTP + PoE
CONNECTIVITYC1 / 5G + Wi-Fi 6
AUTONOMYANT1 / complete AMR
Built to fit your product
Adopt the stack at your layer.
Olive sensor products
Deploy an intelligent ROS-native Olive device as a finished component in your robotic system.
- PRODUCTION-READY HARDWARE
- OLIXOS™ + OLIXAI™ INCLUDED
- NATIVE ROS 2 / DDS
SiP or SoM integration
Bring Olive’s compact compute and software platform inside your own sensor, actuator or robot module.
- 38 × 38 MM PLATFORM
- UP TO 6 TOPS
- FLEXIBLE SENSOR + NETWORK I/O
AMR architecture
Build on the full Olive stack—from embedded electronics and sensors to olixOS™, olixAI™, ROS 2 autonomy and fleet operations.
- MULTIMODAL PERCEPTION
- SLAM + NAV2 + MOTION CONTROL
- ROBOT LIFECYCLE + FLEET
Build with Olive technology
Engineer the whole path. From signal to motion.
Whether you need a board-level embedded platform, an intelligent sensor or a complete autonomous mobile robot stack, Olive can help build the electronics, runtime and robotics system as one architecture.
Discuss your architectureReady to Build Smarter Robots?
Discover how olive’s ROS-native hardware and embedded AI can accelerate your robotics project — from prototype to production.
Backed by Leading Programs & Partners
At Olive Robotics, collaboration drives progress. We are proud to be part of world-class programs, accelerators, and research ecosystems that amplify our impact, accelerate innovation, and extend our reach in the robotics industry.
The information, specifications, and materials provided on this website are published by Olive Robotics GmbH for general informational and commercial purposes only. While every effort is made to ensure accuracy and timeliness, all specifications, performance figures, availability details, and product roadmaps remain non-binding and subject to modification, withdrawal, or correction without prior notice. Nothing on this website shall be construed as constituting a warranted characteristic within the meaning of §§ 434 et seq. BGB, nor as a binding commercial offer within the meaning of § 145 BGB, unless expressly confirmed in writing by Olive Robotics GmbH.
Access to and use of this website is at the user’s sole risk. Olive Robotics GmbH disclaims, to the maximum extent permitted by applicable law, all liability for direct, indirect, incidental, or consequential damages arising from the use or misuse of this website, its content, or any linked external resources. Pursuant to §§ 7–10 TMG, Olive Robotics GmbH is not responsible for third-party content accessible via external hyperlinks and assumes no responsibility for the availability, accuracy, or legality of such content.
All trademarks, logos, product identifiers, and intellectual property referenced on this site, including but not limited to olive®, olixSense™, olixAI™, olixOS™, and related marks, are the exclusive property of Olive Robotics GmbH. Unauthorized reproduction, modification, or commercial use is strictly prohibited. References to third-party technologies or marks are for descriptive purposes only and imply no endorsement or affiliation.
All quotations, orders, and deliveries are governed exclusively by Olive Robotics GmbH’s General Terms and Conditions (AGB) as published on this website. For consumers, statutory rights under the German Civil Code (BGB) remain unaffected.
Olive Robotics GmbH
Embedded Physical AI for Robots.