Autonomous Mobile Robots (AMR) and Automated Guided Vehicles (AGV) are unmanned robots that combine mobility, perception, and connectivity capabilities. They are primarily used for transporting and moving loads of various weights and sizes, as well as other functions. These systems can interact with humans to varying degrees depending on their technology and intended use, operate safely around people, and achieve high levels of cooperation and collaboration. This article will introduce the development and trends of smart robots such as AMRs and AGVs, along with related solutions offered by onsemi.
Smart robots driving the digital transformation of manufacturing in industry 5.0
As the industry approaches Industry 5.0—the next phase of manufacturing digitization—there is a growing demand for enhanced human-machine interaction and robotic functionalities. The application of smart robots ranges from robotic arms to wheeled autonomous delivery robots, and even fully autonomous humanoids, with developments gradually maturing. Unlike traditional industrial robots, intelligent robots employ a variety of sensors, artificial intelligence (AI), and advanced algorithms to interact with their environment, detect obstacles, and collaborate with humans and other machines.
The advantages of deploying autonomous robots include increased productivity and efficiency. They perform repetitive and/or time-consuming tasks, allowing human workers to focus on activities with higher added value. Systems with lighter payloads can be powered by 12V batteries, while higher voltages (such as 48V) can be used to reduce operating currents, thus lowering wiring sizes and costs.
Today, thanks to advancements in artificial intelligence, the global smart robot market is experiencing significant expansion. This enables the construction of more sophisticated autonomous robots that can be deployed not only in warehouses but also outdoors and in less controlled environments. Autonomous robots offer solutions across various industries such as e-commerce, manufacturing, and healthcare and can be highly customized.
Safe battery charger management and voltage conversion systems
onsemi has introduced a comprehensive range of smart and mobile robotic solutions tailored for applications in smart robots. These solutions are highly diverse, and the following will outline the key products based on functional blocks.
First, battery chargers are not part of the robot itself (offboard) and typically involve converting single-phase 120–230 Vac to 12–48 V battery voltage. Chargers can be wired or wireless. AC/DC chargers generally consist of two subsystems: a Power Factor Controller (PFC) stage and a Resonant (LLC) Converter.
In the power factor controller, the NCP1680 is a bridgeless totem-pole CrM (Critical Conduction Mode) PFC controller. It supports constant on-time CrM and valley synchronized frequency foldback, enabling efficiency optimization across the entire load range. It features AC line monitoring, AC phase detection, a novel valley sense scheme, and zero-current detection. It eliminates the need for external components like a Hall sensor for cycle-by-cycle current limiting and is suitable for power levels below 350W. Another power factor controller, the NCP1681, is a bridgeless totem-pole multi-mode PFC controller. It supports fixed-frequency CCM (Continuous Conduction Mode), constant on-time CrM, and valley switching frequency foldback. With proprietary current sense and valley sense schemes, it is suitable for high power, supporting multi-mode applications up to 1kW and CCM applications exceeding 2.5kW.
Additionally, the offline controller NCP4390 is a secondary-side LLC resonant converter with synchronous rectifier control, packaged in SOIC-16. It features dual-edge tracking synchronous rectification control, supports a wide operating frequency range (39kHz to 690kHz), non-zero voltage switching prevents (NZP) through compensation cutback (frequency shift), and allows programmable dead times for primary side and SR switches.
The battery, battery management, and voltage conversion systems constitute the robot's onboard components. Typically, mobile robots use lithium-ion or lithium iron phosphate (LiFePO4) batteries. Lithium-ion batteries are more common due to their higher energy density. On the other hand, LiFePO4 batteries are more stable, less prone to overheating, and have lower nominal voltage. The most commonly used battery systems are 12–48V, which can be paralleled to improve performance.
A typical 24V battery has a capacity of around 50Ah and weighs approximately 10kg. The power tree supplies all logic-level and low-voltage power rails in the system. It usually does not require isolation (battery voltage is below 50V) and works with low dropout regulators (LDOs) to implement multiple parallel buck converters.
onsemi’s FAN65008B converter is a synchronous buck regulator that integrates high-side and low-side power MOSFETs and a fixed-frequency voltage-mode PWM controller. It provides a wide range of voltage conversion with efficiency exceeding 95% at output currents above 2A. It supports a wide input voltage range from 4.5V to 65V and delivers 10A continuous output current. Additionally, it supports programmable switching frequencies from 100kHz to 1MHz and includes features like thermal shutdown, UVLO, overload and short-circuit protection.
Advanced communication and protection features for AMR/AGV
All building blocks in AMR/AGV systems need to communicate with each other. Currently, many communication methods are available. Traditionally, CAN, LIN, RS-485, RS232, and other methods have been used. However, all these communication methods can now be replaced with 10Base-T1S, which allows multiple PHYs to connect to a common bus using just a single twisted pair. This reduces the number of switch ports required and eliminates the need for gateways. 10BASE-T1S requires only an unshielded twisted pair (UTP), significantly lowering cabling costs.
onsemi's Ethernet controller NCN26010 is a 10Mb/s industrial Ethernet MAC+PHY IC controller supporting 10BASE-T1S, compliant with the IEEE 802.3cg standard. It integrates a MAC and 10BASE-T1S PHY and supports PLCA burst mode. If any node needs to send more data than all other nodes, it allows each PLCA transfer opportunity to send more frames. It supports more than eight nodes over 25m of UTP cable, offers enhanced noise immunity, includes a globally unique MAC address, and comes in a 32-pin QFN package.
In terms of smart protection, technologies like eFuse and SmartFET enhance efficiency and reliability while reducing robot downtime. eFuse is a self-protected, resettable electronic fuse commonly used for monitoring input/output voltage, output current, and temperature. The eFuse prevents overcurrent, overvoltage, and high temperature, protecting downstream components, connectors, and PCB traces from damage. It can be used in hot-plug scenarios and situations requiring inrush current limiting. SmartFET is suitable for protecting low-voltage power rails (e.g., 12V) by offering short-circuit protection, inrush current handling, thermal shutdown with automatic restart to prevent high temperatures. It also includes overvoltage protection.
onsemi's current protection solution, the NIS3071, is a four-channel eFuse supporting four independent channels, each up to 2.5A and 60V. It is highly scalable, with outputs combinable to increase the current limit to 10A. It supports thermal protection per channel, digital enable, common fault pins, adjustable turn-on time control, and adjustable overcurrent limit.
The protected MOSFET, NCV84045, is a fully protected single-channel high-side driver with advanced protection features. It supports CMOS-compatible control inputs, delivers output currents up to 32A, and has an extremely low typical RDS(ON) of 50mΩ. It offers diagnostic feedback with current sense output, integrated clamp for inductive switching, loss of ground and loss of VD protection, as well as ESD protection and short-circuit protection.
Complete motor control and position sensing solutions
The central processing unit (CPU) serves as the "brain" of the entire system, managing all internal system communications as well as interactions with the outside environment. Depending on the complexity of the system, it must have sufficient computational capability. Additionally, smart robots require support for Simultaneous Localization and Mapping (SLAM), a method for creating maps of unknown environments. Mobile robots use SLAM algorithms to navigate autonomously in their surroundings. On the other hand, motion and actuator control are used to manage the wheels that move the robot. Robots can also employ robotic arms or lifting mechanisms to handle loads. All these functions typically rely on brushless DC (BLDC) motors, which require sophisticated algorithms for precise control.
onsemi’s gate driver NCD83591 is a three-phase gate driver ideal for industrial applications. It supports an operating power supply range of 5-60V and provides up to 250mA FET constant current drive. It can execute motor PWM up to 30kHz and features individual six-gate control mode. It integrates protections such as UVLO, HBM and CDM ESD, internal gate pull-down during power loss, and more, and comes in a 28-pin QFN package.
The MOSFET NTMJST2D6N08H offers 2.8mΩ RDS(ON) and 80V VDS, with low capacitance and gate charge to reduce switching losses. It comes in a TCPAK57 package, which dissipates heat from the top, lowering PCB temperatures and improving PCB utilization.
The MOSFET NTMFS0D4N04XM is a best-in-class 40V MOSFET from the T10M family, designed for BLDC motor drives. It supports 0.42mΩ RDS(ON) to minimize conduction losses and comes in a 5mm x 6mm package. It features excellent soften recovery characteristics, and its low voltage spikes reduce stress and EMI issues.
Inductive position sensors measure the rotation of wheels or other moving parts to accurately track their position and orientation in the environment. They can be part of the electronic commutation for BLDC control. Inductive encoders offer many advantages over traditional optical or magnetic sensing, being robust, lightweight, requiring few components, and resistant to vibration or contamination.
Inductive sensing solution NCS32100 is a contactless sensor system comprising two PCBs: one rotor with two printed inductors (no soldered components) and one stator with printed inductors and an encoder IC. Compared to traditional optical encoder solutions, which may require over 100 components to function, the NCS32100 requires only 12 components to achieve minimal functional systems.
onsemi’s NCS32100 calculates position and velocity and features an absolute encoder, capable of determining its position without movement. It supports full accuracy at 6,000 RPM (up to 45,000 RPM). A 38mm sensor provides an accuracy of ±50 arcseconds (0.0138 degrees) or higher. It can differentiate and reject vibrations caused by rotational movement, supports 20-bit single-turn and 24-bit multi-turn resolution output, integrates a CortexM0+ MCU, and is highly configurable. It is a more affordable alternative to various optical encoders, with a self-calibration function executable through a single command.
Advanced lighting, sensor, and wireless solutions
The robot can be equipped with LED lights to indicate their presence to people and other robots. Compared to older light sources, LEDs are more efficient, lighter, and longer lifespan. Depending on the color of the light, robots can indicate directions, signal statuses, and more. Based on the required power and application, onsemi offers a variety of LED drivers and controllers.
The linear LED driver NCV7685 is a 12-channel LED linear current driver with I2C support, capable of driving 12 parallel constant current channels, each providing 60mA sink current. It supports 128 levels of duty cycle, with each channel independently controlled. It includes advanced diagnostic options and complies with the AEC-Q100 standard.
To enable truly autonomous smart robotic systems that do not rely on a set path (as required by AGVs), the system must prevent collisions with obstacles or other robots, and most importantly, avoid collisions with humans. Depth sensing can be achieved using various sensors such as LiDAR, imaging, radar, and ultrasonic, each with its strengths and weaknesses.
Combining data from multiple sensor modalities is known as sensor fusion. Sensor fusion leverages the strengths of all sensors, as no single sensor technology can provide reliable information under all conditions. Multiple operational sensors can work together to provide a more reliable dataset.
onsemi's silicon photomultipliers array RDM-0112A20 is an NIR-enhanced SiPM designed for LiDAR applications. It features a 12 SiPM pixel array with a common anode, no fast output, a 20x20 µm microcell active area, and a photon detection efficiency (PDE) of 16% at 905nm. The micro-lens technology enables maximum optical efficiency, with a recommended Vop of 30V.
The silicon photomultipliers MicroFC-100 is a 1x1 mm SiPM suitable for single-point or 2D LiDAR. It supports standard and fast output and offers the highest sensitivity in the visible range. Depending on the microcell size, it has a PDE greater than 18% at 420nm, with a recommended Vop of 25.2V, and comes in 10, 20, or 35 µm microcell sizes.
Optical sensors can be used for depth sensing, orientation sensing, or to provide additional features for robots, such as inspection or image recognition. Different subsystems in robots may contain multiple image sensors (IS) and IS processors. These are the only sensorics solutions capable of detect the colors. Using optical sensors, the system can detect the obstacles, improve safety, or readout information (e.g., barcode data).
onsemi’s image sensor AR0234 is a 1/2.6” 2.3Mp CMOS digital image sensor with a 1920x1200 active pixel array. It features industry-leading global shutter efficiency, excellent low-light and infrared performance, and supports auto exposure, windowing, and row/column skip modes.
The AR0822 is a 1/1.8” 8Mp CMOS digital image sensor with a 3840x2160 active pixel array. It supports rolling shutter, onboard eHDR, high sensitivity, low read noise, intelligent linearization to mitigate motion artifacts and LED flicker, and enhanced NIR response.
Wireless connectivity systems can be used to collect sensor data, monitoring, and localization, as part of SLAM algorithms. Bluetooth® Angle of Departure (AoD) is an indoor positioning system that functions similarly to outdoor GPS. In the AoD method, the transmitting device uses multiple antennas arranged in an array to transmit 250kHz signal. The receiving device has one antenna and collects data as the signal from the transmitting device passes through it, enabling the calculation of the signal's direction.
onsemi’s Bluetooth Low Energy RSL15 is a wireless microcontroller unit (MCU) that adopts Bluetooth 5.2 technology and a secure Arm® Cortex®-M33 processor. With built-in power management, flexible GPIO and clocking schemes, and a wide power supply voltage range, the device provides maximum design flexibility for high-performance and ultra-low-power applications. It optimizes energy efficiency, minimizes battery drain, reduces battery size, and extends the lifespan of battery-powered sensors. Equipped with an industry-leading ultra-low-power microcontroller and an easy-to-use SDK, it is the industry’s lowest-power flash-based secure Bluetooth® Low Energy MCU.
Conclusion
The development of smart robots is advancing at an unprecedented pace. From industrial production to household services, and even in medical and educational fields, the application scenarios of robotic technology are becoming increasingly diverse. With the continuous progress of artificial intelligence, the Internet of Things, and sensing technologies, smart robots are not only becoming more efficient, precise, and versatile but are also better aligned with human needs. onsemi offers a comprehensive range of product solutions to accelerate the development of smart robots, helping to build a more efficient, convenient, and intelligent future world.