That takes a multi-layered, real-time awareness of their surroundings to operate safely and effectively alongside humans, including walking and grasping objects. Data from multiple sensor modalities like cameras, mm-wave radar, LiDAR, IMUs, encoders, force sensors, and more must be fused and integrated into a single, cohesive 3D model.

Sensor fusion relies on a variety of software tools to integrate multimodal sensor data into a single unified spatial model. That provides humanoids with the situational awareness required to complete complex human-like tasks like navigation, walking, and grasping objects in real-time.

It also supports safe and predictable interaction with nearby people and other robots. From the bottoms of feet to fingertips and shoulders, data from tactile force, torque, and load sensors help track and maintain dynamic physical interactions (Figure 1).

Being aware

Situational awareness is not an end; it’s the first part of a continuum of perception and participation by humanoids in their surroundings. It relies on a three-layer architecture and enables humanoids to independently perform useful tasks.

The process begins with perception, the humanoid’s senses, and the sensors. Masses of heterogeneous data about the environment are consumed from 3D LiDAR, cameras, tactile force, microphones, IMUs, and more. That data is fused to construct a 3D digital map, and as the basis for understanding and interacting with people.

That information is combined with the requirements of specific tasks like walking or grasping objects in the reasoning phase using a variety of AI/ML tools. The reasoning phase develops detailed task plans to control humanoid motion and interactions with its surroundings. The results of reasoning are translated into task execution using physical AI (PAI) tools.

Circular process

Situational awareness is a circular process. As tasks are executed, physical states of elements like feet, legs, arms, hands, and so on (proprioception) are combined with external environmental data (exteroception) to enable humanoids to adapt in real time to unexpected developments.

That adaptability is required to ensure that the results match the expectations generated by the reasoning and task planning step. That demands continuous and complex real-time sensor fusion.

Three levels of fusion

It takes multiple levels to effectively fuse data from multiple sensor modalities (Figure 2). Fusion starts with raw data. This data-level fusion uses techniques like Kalman and complementary filters, weighted averages, and other tools to diminish noise and improve data quality. Features are also extracted at this level.

Feature fusion is the second level and combines the features into a unified feature vector structured to provide a more comprehensive picture of the overall sensor data. Commonly employed tools include factor analysis and principal component analysis (PCA), along with multidimensional scaling (MDS) or other dimensionality reduction methods to increase the efficiency and effectiveness of the classification process.

Decision-level fusion is especially important when accuracy is crsitical but there is limited training data or high uncertainty. Methods like majority voting and Bayesian networks are used to perform a combined analysis on predictions or classifications from several models to arrive at an improved result.

Morphological computations 

The use of hands, in place of less capable end-effectors, is a key for deployment of advanced humanoids. Wrist and finger touch (tactile force) sensors can be fused with visual information from RGB cameras to identify and safely grasp and manipulate even fragile objects. Development of simplified methods is being pursued.

For example, a prototype was fabricated using multi-material 3D printing. Compliant materials were used for the sensors and joints, but a rigid material was used for the “bones” to provide support.

The use of compliant materials means the hand conforms to the shape of an object without the need for positional feedback. The use of soft sensors further ensures more delicate contact with objects (Figure 3).

The tactile data acquired from grasping objects is analyzed using morphological computation, where the system’s physical form and dynamics perform tasks without the need for energy-intensive external processing. In this prototype system, a multiple-grasp implementation was used to enhance object identification.

Summary

Situational awareness is needed to support the effective and safe operation of humanoids. It requires the fusion of data from diverse sensor modalities, AI/ML analysis of the resulting data, and translation into execution of specific tasks. It’s a circular real-time discipline that’s still being refined.

References

Beyond Wheels: Designing and Building a Walking/Bipedal Robot, KEYi Technology
Energy-Aware Sensor Fusion Architecture for Autonomous Channel Robot Navigation in Constrained Environments, MDPI sensors
Enhancing Human–Robot Collaboration through a Multi-Module Interaction Framework with Sensor Fusion: Object Recognition, Verbal Communication, User of Interest Detection, Gesture and Gaze Recognition, MDPI sensors
Humanoid Factors: Design Principles for AI Humanoids in Human Worlds, arXiv
Humanoid Robot Reference Guide, STMicroelectronics
Multi-Sensor Data Fusion and CNN-LSTM Model for Human Activity Recognition System, MDPI sensors
Recent Advancements in Humanoid Robot Heads: Mechanics, Perception, and Computational Systems, MDPI biomimetics
Safety meets speeds: how next-gen sensors free robots from cages, imec
Soft Robotic Hand with Tactile Sensing, Oxford Robotics Institute
The Humanoid Robotics Nervous System: Perception, Balance, and Real-Time Control, Panasonic
What Is Morphological Computation? On How the Body Contributes to Cognition and Control, MIT

Related EEWorld Online content

What kinds of PAI dev kits are available for humanoid robotics?
How do humanoid robots and physical artificial intelligence fit into industry 5.0?
Does Industry 5.0 exist and what does it mean?
How is power limiting the adoption of physical artificial intelligence in humanoid robotics?
What is physical artificial intelligence and why is it important?

The post Humanoid robots, sensor fusion, and situational awareness appeared first on Sensor Tips.

STMicroelectronics has introduced ST SafeSense VD56GA, a 1.1 MP automotive image sensor for infrared driver and occupant monitoring. Housed in a 3.7 mm × 3.2 mm CSP package, the sensor includes embedded image-processing functions, supports RAW and YUV outputs and uses the ST DeepNIR pixel architecture for infrared imaging. Its integrated functions include mirror, crop, dark calibration, auto exposure and piecewise-linear processing to reduce the need for external image-processing resources.

The post Compact image sensor integrates embedded processing functions appeared first on Sensor Tips.

This article examines how zonal architectures relocate sensors, consolidate the wiring harness, affect communication latency, and reshape fail-operational redundancy.

How does zonal design change sensor placement?

Conventional vehicles use a domain-based in-vehicle network architecture (DIA). Electronic control units and sensors are grouped by function, such as powertrain, infotainment, body, and so on, then wired across the vehicle to a central domain gateway. That arrangement worked while each function kept its own controller.

Rising cross-domain features and sensor counts have turned the harness into one of the heaviest and most complex systems in the car. A zonal-based in-vehicle network architecture (ZIA) groups components by physical location instead. Let’s look at the same vehicle drawn two ways, as shown in Figure 1.

The geometrical view splits the vehicle into four physical zones: front-left, front-right, rear-left, and rear-right. The functional view shows the in-zone components, a battery, a switch, and a sensor ECU, sharing energy and data connections. This platform predates fully comprehensive zone controllers, so it spreads those roles across separate units.

A modern ZIA folds these roles into one zone control unit (ZCU). The ZCU becomes a local hub for power distribution and data handling. External sensors, radar, lidar, and cameras, for example, connect to the nearest ZCU. They no longer run all the way back to a functional controller. Placement now follows the chassis geometry, which makes the wiring savings possible.

How much wiring does zonal consolidation save?

Connecting sensors to the nearest zone removes the long dedicated runs back to a central domain controller. The benefit scales with the number of zones and the number of ECUs, as shown in Figure 2.

At 120 ECUs, a 10-zone ZIA cuts total harness length by 63.7% compared to a five-domain DIA. The reduction grows as zones increase, since more ECUs reach their hub through shorter local runs. An eight-zone layout reaches about 61% at the same ECU count.

Independent estimates put the harness weight reduction from zonal design at 15 to 20%, which frees packaging space in dense autonomous platforms. The savings carry a cost. Moving heavy processing off the distributed ECUs forces the architecture to lean on a centralized high-performance computing unit (HPCU) that acts as the backbone gateway. Less copper in the harness means more dependence on one central brain.

How does zonal design affect safety-critical latency?

Shorter wiring is a physical win. The data path raises a separate question. High-priority command-and-control signals must meet strict end-to-end (E2E) delay limits to actuate safely. Figure 3 compares E2E delays for the highest-priority scheduled traffic in both architectures.

The E2E delay for the first control data stream drops from 36.68 µs in the DIA to 19.52 µs in the ZIA. Vehicle-to-everything (V2X) traffic falls from 31.51 to 14.34 µs. Five of the seven traffic types are faster under the ZIA, because data crosses fewer switches and links on the backbone.

The gain is not universal. Navigation data runs slower in the ZIA, rising from 9.16 to 17.74 µs, while lidar data stays effectively unchanged.

What does fail-operational redundancy cost in a zonal system?

Autonomous platforms have no human driver to fall back on. Safety-critical ECUs must move from fail-safe to fail-operational behavior. Consolidating functions into a few zone controllers widens the impact of any single fault.

Meeting ISO 26262 Automotive Safety Integrity Level D (ASIL D) requires redundancy designed in from the start. The system-level cost of that redundancy shows up in Figure 4.

Figure 4 evaluates four options, domain and zonal layouts, each with vehicle-centralized or controller-based processing. Redundancy is added through logical splitter and merger nodes that route data to separate branches. A single failed controller can take several consolidated functions down at once.

The framework adds redundancy only to selected nodes, such as low-level speed control, then re-evaluates the full architecture. The controller-based zonal mapping holds the best cable length, but adding redundancy raises hardware cost and backbone communication load.

When redundancy is required, the domain controller-based mapping offers the best overall balance across the measured parameters. Zonal layouts win on shorter cable length but carry a higher communication load. That split is why the field is converging on hybrid designs, zones for body and comfort functions, and separate domains for safety-critical control. For safety-critical control, pure zonal is not the right answer yet.

Summary

Zonal architecture reshapes sensor redundancy from the wiring up. Geometric placement shortens the harness, with reductions above 60% at high ECU counts. It also lowers latency for most safety-critical traffic, as long as the zone count stays controlled.

The redundancy picture is less one-sided. Fail-operational safety adds cost and communication load, and the measured data still favors domain controller-based balance. For teams committed to a zonal layout, controller-based zonal mapping is the rational compromise, and hybrid zone and domain designs are where the industry appears to be heading.

References

• Design of Zonal E/E Architectures in Vehicles Using a Coupled Approach of k-Means Clustering and Dijkstra’s Algorithm, Energies, MDPI
• Performance Evaluation of Zone-Based In-Vehicle Network Architecture for Autonomous Vehicles, Sensors, MDPI
• Automotive Architecture Topologies: Analysis for Safety-Critical Autonomous Vehicle Applications, IEEE Access
• Analysis of E2E Delay and Wiring Harness in In-Vehicle Network with Zonal Architecture, Sensors, MDPI

EEWorld Online related content

• Automotive sensor requirements for software-defined vehicles: latency, resolution, and zonal architecture
• How does ISO 26262 (Road Vehicles – Functional Safety Standards) apply to EVs?
• How do EVs benefit from using zonal architectures?
• Zonal E/E will transform automotive architectures
• Why EVs are transitioning to zonal architectures
• What is a sustainable wiring harness?

The post How are zonal architectures reshaping sensor redundancy strategies? appeared first on Sensor Tips.

Teledyne e2v has introduced the Flash 1K2 and Flash 8K CMOS image sensors, which deliver 3,000 frames per second at 1K × 1K resolution and 1,800 frames per second at 8K × 1K resolution for high-speed inspection and 3D laser profiling. Both sensors use 6 µm global shutter pixels and support HDR modes up to 100 dB, multiple regions of interest and frame-to-frame programmability. The single-tap architecture supports integration across the Flash family, which also includes 2K × 1K and 4K × 1K models with a common feature set and development platform.

The post Global shutter sensors support high-speed inspection appeared first on Sensor Tips.

Sensata Technologies has introduced its R290 Pressure + Temperature (P+T) Sensor for electric vehicle thermal management systems using R290 (propane) refrigerant. The sensor combines pressure and temperature measurement in one device, with a leakage-optimized design and construction intended for automotive operating conditions. It supports air-conditioning and thermal management systems in battery electric vehicles, plug-in hybrid electric vehicles and other electrified platforms adopting low-GWP and PFAS-free refrigerants.

The post Combined sensing simplifies electric vehicle thermal management appeared first on Sensor Tips.

Bourns, Inc. has expanded its Bourns® PEC11S series 12 mm surface mount incremental encoders with additional shaft lengths and SPST momentary switch options for compact control interfaces. The contacting encoder provides 2-bit quadrature output, up to 60 rpm operating speed, 3 Ω maximum closed-circuit resistance and a -40 °C to +70 °C operating range, with minimum life ratings of 15,000 rotational cycles and 20,000 switch cycles. The added mechanical options help engineers simplify panel layout, reduce redesign work and combine rotary and push input functions in one component for industrial, consumer and communications equipment.

The post 12 mm encoder adds shaft and switch options appeared first on Sensor Tips.

Where temperature sensors are placed in a pack

Temperature sensors provide primary feedback to the BMS, feeding cell and module temperature data into algorithms that set charge current limits, actuate cooling systems, and enforce safety interlocks. Engineers deploy these sensors across cells, modules, cooling hardware, and pack boundaries. This placement helps the system correlate electrical stress with local thermal behavior.

As shown in Figure 1, temperature sensors operate within a broader BMS feedback path that combines voltage, current, and thermal data to support control, diagnostics, and safety functions.

At the cell level, engineers mount sensors at or near tabs, along the sidewall at mid-height, or between adjacent cells. These locations help capture current-induced heating, cell-to-cell gradients, and thermal paths from the cell interior.

As shown in Figure 2, cell-level temperature sensing may use embedded sensors in pouch cells or surface-mounted thermocouples on cylindrical cells, depending on cell format and the measurement point the BMS needs to represent.

Module-level designs typically use only a few sensors, often at the coolant inlet, coolant outlet, or in a “penthouse” position above the cells. This configuration captures average temperature and coarse gradients, yet inherently undersamples individual cell behavior. As a result, localized cell heating may remain undetected until it affects the module-level reading.

Cooling hardware and enclosure sensors provide complementary visibility into heat rejection and pack-level boundary conditions. Sensors on cooling plates, manifolds, and heat sinks indicate coolant temperature rise, flow uniformity, and heat extraction performance. Enclosure or case-level sensors add boundary data and help detect global overheating driven by ambient conditions, underperforming cooling circuits, or a propagating thermal runaway event.

How placement determines what a sensor reads

A temperature sensor should approximate the cell or module temperature that the BMS uses for control rather than overrepresenting the local mounting interface. Placement determines whether the reading tracks cell-level risk or is biased by local boundary conditions, including coolant flow, tab conduction, sensor contact quality, and insulation layers.

These biases are most consequential around heat sinks and conductive paths. High coolant flow or a highly conductive tab may cause the sensor to read cooler than the cell core, underestimating peak internal temperature and masking core-to-surface gradients.

Thermal coupling determines how quickly and accurately the sensor follows the monitored surface. A sensor pressed against the cell case or tab reads a temperature filtered by material conductivity, contact quality, and any intervening insulation. A sensor loosely taped to an outer wrap responds slowly and attenuates the temperature gradient it is intended to capture, degrading hot-spot localization.

Boundary and cell-oriented measurements support different control functions. Coolant inlet and outlet sensors capture system-level heat rejection yet miss local defects such as a blocked channel or poorly seated cell. Sensors placed closer to bottlenecks, including coolant headers or points within the cooling plate, expose flow irregularities that inlet and outlet measurements cannot detect.

For cell-oriented sensing, engineers often mount sensors midway up the cell sidewall. At this location, conduction paths from the tabs and cell interior converge, providing a practical compromise between core-temperature representation, assembly access, and mechanical security.

Sensor density and hot-spot detection

Sensor count improves observability only when paired with targeted placement. Low-density schemes, with one or two sensors per module, report average temperature and a coarse gradient across many cells. A single underperforming cell or local contact defect may remain undetected until it produces enough heat to shift the module-level average.

Advanced distributed sensing increases spatial resolution. Temperature-monitoring tape and fiber-optic arrays along cell rows and cooling channels support near-continuous temperature maps instead of sparse point measurements. This added resolution can reveal small hot spots, flow maldistribution, and uneven clamping before the resulting temperature rise appears in a pack-level average. It also helps differentiate genuine anomalies from normal thermal variation.

As shown in Figure 3, optical fiber sensing can place measurement points at both external and internal cell regions, extending thermal visibility beyond sparse surface measurements.

Adding more sensors does not automatically improve detection. Thermal sensor allocation research in other thermally dense electronic systems shows that targeted placement in high-sensitivity regions can reduce maximum hot-spot estimation error more than uniform spacing with the same sensor count. The same principle applies to EV packs: extra sensors may add little unless thermal coupling and model fidelity also improve.

Effect on BMS control and safety

Sensor observability determines how closely the BMS can operate the pack to its thermal limits. Well-placed, well-coupled sensors provide enough spatial resolution and measurement fidelity for the BMS to adjust current limits by module or zone, vary coolant flow, and command pumps and valves while maintaining cell temperature margins.

Sparse, poorly located, or weakly thermally coupled sensors push the BMS toward conservative current limits because the system must account for conditions the sensors cannot rule out.

The same observability limit applies to detection speed. A localized temperature rise, evaluated alongside voltage and current data, helps flag abnormal cell behavior and identify thermal runaway risk. Sparse sensing delays detection and increases false-negative risk.

To estimate temperatures between measured points, many BMS designs pair sensor data with predictive thermal models, including lumped or distributed thermal networks. These models estimate core temperatures at unmeasured locations and project temperature trajectories during fast charging or high-power operation.

When sensor placement and density match the model’s assumptions, the combined system can separate benign thermal transients from genuine faults, reducing unnecessary derating or shutdowns while maintaining safety margins.

Design and integration considerations

Sensor placement decisions usually begin with thermal modeling. Engineers run computational fluid dynamics (CFD) and finite element analysis (FEA) simulations early in the design process to identify high-risk regions, including end cells, pack corners, coolant inlet and outlet areas, and structural interfaces. They then specify sensor locations that maximize hot-spot observability within the available sensor count.

Mechanical and manufacturability constraints narrow these options. Sensor placement must account for assembly sequence, isolation, creepage, and clearance requirements, while avoiding crush zones and structural load paths. Peel-and-stick thermistors are commonly mounted midway up the cell, where the assembly process is less likely to shear them. Bolt-on sensors may be placed on terminals when the design allows direct access.

Contact quality determines whether the placement strategy performs in production. Thermal tape, silicone adhesive, or clamping features keep the sensor pressed against the cell or cooling plate so the reading tracks the monitored surface.

Inadequate contact adds thermal resistance and biases the reading toward ambient temperature. Pressure-and-temperature mapping film systems are increasingly used during development and manufacturing to verify intended thermal contact and refine sensor placement before production starts.

Summary

Sensor placement defines what a temperature reading represents: cell-level risk, a coolant boundary, or an averaged module condition. Sensor density determines whether the BMS can resolve a localized hot spot or track only a pack-wide trend. Moderate sensor density targeted to thermal risk can outperform higher sensor counts that ignore placement quality. Together, placement and density set the practical limits on BMS control fidelity and fault detection speed.

References

• 7 Places for Electric Car Temperature Sensors in Vehicle Design, Amphenol
• A Manufacturer’s Guide to EV Thermal Management & Sensor Technology, Amphenol
• What Role Do Temperature Sensors Play in EVs and Battery Management, Coherent Market Insights
• Sensing-Based Monitoring Systems for EV Battery – A Review, ScienceDirect
• Advancements in Thermal Management and Safety of Li-ion Batteries for EVs, ScienceDirect
• Research of Thermal Sensor Allocation and Placement Based on Dual Clustering for MPUs, NIH
• How Temperature Sensors are used in Battery Pack Temperature Management for EVs, TC-Inc
• Fiber Optic Temperature Monitoring for EV Battery, XRAY
• Beyond the Battery Pack – How Temperature Mapping Can Be Used Across the Electric Vehicle, TekScan
• Integrating Battery Models into BMS Workflows, NeoVarsity

Related EE World content

• How to Ensure EV Battery Safety with Advanced Temperature Monitoring
• AI-Enhanced Thermal Imaging Identifies Early Signs of Thermal Runaway
• When EV Thermal Management Becomes a System Problem
• How do Increased Voltage Levels and Faster Switching Speeds Influence Sensor Placement and Accuracy in EV Systems?
• How Many Thermal Sensors are There in an EV?

The post How sensor placement and density impact thermal mapping in EV battery packs appeared first on Sensor Tips.

Texas Instruments (TI) has introduced the TMCS2100-Q1, a multiaxial coreless Hall-effect current sensor for HEV and EV traction inverters. The device measures magnetic fields on horizontal and vertical axes and is specified for displacement error below 1% at 0.4 mm movement and as low as 0.25% at 0.1 mm, which supports tighter inverter current control under vibration and changing load conditions. The sensor architecture does not require busbar notches or holes, helping engineers reduce inverter size and simplify mechanical integration in 800V and other high-voltage EV platforms.

The post Dual-axis current sensor cuts displacement error in EV inverters appeared first on Sensor Tips.

Vishay Intertechnology, Inc. has introduced the Vishay Semiconductors VCNL36829UM, a fully integrated short-distance proximity sensor in a 1.6 mm x 1.0 mm x 0.35 mm transparent mold package with VCSEL, photodiode, ASIC, 16-bit ADC and smart dual I²C slave address. The sensor supports a 50 mm operating range, 14-bit proximity resolution, sunlight immunity up to 200 klx, 5 μA idle current and a 1.65 V to 2.0 V supply range for battery-powered and space-constrained consumer designs. Applications include wear detection in TWS earphones, VR / AR headsets and smart glasses as well as touchless dispensing and lid-off detection in consumer appliances.

The post Proximity sensor integrates VCSEL, photodiode and 16-bit ADC appeared first on Sensor Tips.

Microchip Technology has released Revision 2.0 of its PolarFire® FPGA Ethernet Sensor Bridge, an HSB-enabled board for standardized sensor integration using NVIDIA Holoscan Sensor Bridge technology. The board supports up to four cameras, 10Gb Ethernet-based architectures, USB-C power, MIPI® CSI-2®, I²C®, UART and GPIO interfaces, with expansion support for SLVS-EC 2.0, 12G-SDI, HDMI® and DisplayPort. It is designed for edge AI systems in medical, industrial and humanoid robotics applications using NVIDIA Jetson and IGX platforms, helping developers reduce multi-interface sensor connections, validate latency and support compact deployments.

The post Sensor bridge supports four cameras for edge AI appeared first on Sensor Tips.

Danisense has introduced the DP12IP current transducer for isolated DC and AC current measurement in PCB-mounted power electronics. Based on closed-loop compensated fluxgate technology, the DP12IP measures up to 18 A with a maximum linearity error of 10 ppm and supports measurement resistors up to 100 Ω at full scale. The 32-mm, 250-g transducer is designed for 1U power supplies, precision servo drives, battery test benches, battery emulators, magnet power supplies, power measurement systems and power analysis systems.

The post Current transducer measures DC and AC up to 18 A appeared first on Sensor Tips.

Magnetic Sensor Systems has introduced the R-10-200-CWM Series clockwise, forward shaft extension rotary solenoids with 18 AWG winding configurations from 20 to 37 AWG. The 2.00-in.-diameter solenoids provide 40° of clockwise rotation, support 100% to 10% duty-cycle options and deliver starting torque from 11 oz-in. to 49 oz-in., depending on winding selection. The Made in USA solenoids include an integral return spring, dual radial ball bearings and mounting options for use in medical devices, automated assembly systems, packaging machinery, security interlocks and other OEM equipment.

The post Rotary solenoids offer 18 AWG winding options appeared first on Sensor Tips.

Sensata Technologies has introduced the Active + Passive PyroFuse, the STPS500P Series, for high-voltage protection in electric vehicles, commercial transportation, charging infrastructure and industrial electrification systems. The device combines signal-triggered pyrotechnic interruption with a mechanically driven current-based passive trigger in a single package, providing redundant protection and millisecond-level response independent of current level to help protect contactors, busbars and downstream components. By combining active and passive protection in one device, the PyroFuse can help OEMs simplify high-voltage architectures, reduce component sizing and lower overall system cost and complexity.

The post Dual-trigger fuse adds redundant HV protection appeared first on Sensor Tips.

The increasing number of sensors in modern systems and the diversity of sensor types across applications like machine learning (ML) and automation present an expanding attack surface that can be invaded by bad actors.

Sensors are used in a variety of critical systems, from the electric grid to autonomous vehicles and pacemakers. Corrupted data can create life-threatening conditions. Corrupted data can also result in an AI application that makes incorrect predictions, resulting in unsafe actions. Adding to the challenges, sensors are often resource-constrained devices, making it difficult to integrate significant levels of security directly into the sensor.

The growing complexity of sensor networks further expands the attack surface, making implementing sensor security complex as well as critical. That necessitates implementing a defense-in-depth approach that embraces hardware, network, and software layers (Figure 1).

Strategies for sensor security

Hardware considerations for sensor security include tamper detection using physical switches, accelerometers, or other tools to monitor for unauthorized physical access. Unused ports should be physically sealed and secure boot implemented.

Network isolation can be important, including the use of virtual LANs to limit outbound sensor communication and block all unauthorized inbound connections. Use of a zero-trust architecture to verify all communication provides an additional level of security.

A well-regulated and controlled patch management process for the delivery of encrypted firmware updates is essential. ML tools can be used to identify anomalous sensor data that may indicate a sensor that’s been physically compromised or subjected to environmental manipulation.

Holistic approach

Strategies for sensor security must extend beyond the edges of traditional networking. Modern systems no longer include a so-called ‘air gap’ without wired or wireless connections that isolates individual systems from the rest of the operation. Today, most systems are designed to allow various types of external connectivity, adding dimensions of concern to the attack surface (Figure 2).

• Remote virtual private networks (VPNs) used for remote and global connectivity can be hacked.
• Manufacturing execution systems (MES) link high-level business planning systems (like enterprise resource planning, ERP) and the physical production floor.
• Engineering laptops and USB drives are often used to update systems and backup configuration data.

Wired vs. wireless

There are fundamental differences in attack surfaces and attack vectors between wired and wireless sensor implementations. Wireless connections cannot be disabled by simply cutting a wire. Wireless sensors can be targets for signal interception or manipulation.

Wired sensor networks are less flexible than wireless implementations and are vulnerable to communication or power cables being severed. Insertion of a resistor at the sensor end of a connection can allow the control panel to detect if a wire has been cut, or if a sensor has been tampered with. Wired connections can’t be easily intercepted and are relatively immune to hacking, jamming, and electromagnetic interference.  

OT vs IT

Finally, there’s a tension between the security demands of operational technology (OT) on the factory floor and information technology (IT) in businesses. For example, OT systems prize stability and infrequent changes while IT systems require more frequent updates to ensure maximum performance.

The intersection between IT and OT systems must be tightly managed. A strictly IT-related event is not generally life-threatening. An OT-related event can compromise safety. OT concerns extend to supply chain issues and ensuring that new sensors or other assets and maintenance or calibration tools don’t introduce security risks (Figure 3).

Summary

Implementation of sensor security in connected systems is as complex as it is important. There are multiple types of sensors, heterogeneous IoT architectures including wired and wireless devices, and application requirements to consider, plus the intersection of IT and OT systems to manage. External vulnerabilities from new equipment and calibration services can exacerbate the internal networking challenges.

References

A step-by-step guide to achieving fast, secure IoT connectivity and device lifecycle management, Crypto Quantique
Combining Edge Computing-Assisted Internet of Things Security with Artificial Intelligence: Applications, Challenges, and Opportunities, MDPI applied sciences
Enhancing Cyber Security Through Machine Learning-Based Anomaly Detection, International Journal of Engineering Research & Technology
IoT Security: Essential Strategies to Protect Connected Devices, Paessler
OT Cybersecurity: The Guide to Securing Industrial Systems, TXOne Networks
Secure OTA Boot Chains and Firmware Verification: Building Trust in Connected Devices, Promwad
Securing All Your Shiny New Connected Devices, NJCCIC
Securing Connected Devices: Enhancing IoT Security in Manufacturing, Orlan Tech
Security Considerations for Field Equipment in Industrial Systems, Industrial Cyber
Security for Industrial Control Systems (ICS), Splunk
Sensors Cybersecurity, MDPI sensors
What Is ICS Security?, Palo Alto Networks

Related EEWorld Online content

How does the Machinery Regulation (EU) 2023/1230 affect designs?
What is an AI governor and how does it relate to physical AI?
How will the Cyber Resilience Act impact embedded developers?
How does the Zenoh protocol enhance edge device operation?
OpenClaw is open-source edge AI for (almost) every application

The post How to implement sensor security in connected systems appeared first on Sensor Tips.

Melexis has introduced the MLX91229, a conventional-Hall current sensor with a sigma-delta digital output for current sensing from 200 A to 2000 A in electrically noisy automotive environments. Designed for traction inverters and other high-power vehicle systems, the IC shifts signal transmission from analog to digital to help maintain signal integrity where EMI, longer PCB traces and wiring can affect measurement accuracy. The device also uses the same footprint as existing analog Hall-effect solutions, allowing evaluation of digital current sensing without major layout changes.

The post Hall current sensor adds sigma-delta digital output appeared first on Sensor Tips.