Bench Talk

Autonomous Robots Meet Consumer Luggage

Fri, 28 Oct 2022 00:27:59 GMT

How Engineers Tackle Consumer Autonomous Robot System Challenges

(Source: scharfsinn86 - stock.adobe.com)

When people hear “autonomous mobile robots (AMRs)”, many imagine relatively large robotic systems operating in order fulfillment centers, warehouses, or various industrial applications. Few would imagine artificial intelligence/machine learning (AI/ML) powered rolling suitcases dutifully following their owners as they travel to destinations across the globe. Even fewer would imagine this technology is available commercially today. The dawn of household AI/ML systems and robotics is here. To further enable this revolution, engineers must overcome several hardware and software hurdles to meet TSA approval and fit all necessary electronic systems within a compact space that doesn't substantially increase weight while keeping costs competitive.

Here, we discuss the trends in small consumer AMR systems—specifically luggage and the various hardware and software concerns—and design challenges associated with bringing small consumer AMR systems to the masses.

AMR Luggage for the 21st Century Traveler

Hauling luggage through airports, train stations, and bus stations is generally one of most people's greatest drags associated with long-distance travel. Hence, it is surprising that there has been little innovation in luggage design over the past few decades. While luggage has undergone an aesthetic overhaul in the past 20 years, not much has been done regarding real technological advancement. Fortunately, this has all changed recently. AMR luggage systems designed to track and follow their owner using only onboard systems are now commercially available.

This is a significant evolution from typical smart luggage, which includes luggage systems that have built-in batteries for charging smartphones, GPS, open detection, and some you can even ride on. With the latest AMR luggage systems, AI/ML technology is used to navigate the dizzying and cluttered environment from home, through a transit station, and on to the user's destination. This type of navigation requires sophisticated vision systems, proximity detection systems, smartphone interfacing, and mobile robotic electronics.

Challenges with Small Consumer AMR System Design

Unlike industrial AMR systems, consumer AMR luggage is relatively small and must remain low-weight and easy to use, even when not operating. Moreover, industrial robots are designed to be safe in intrinsically dangerous industrial environments, where consumer AMR must work in less-controlled environments. Therefore, along with the general need for mobile robots to be efficient, reliable, and responsive, small/personal AMR luggage systems must also account for the chaos of busy transit terminals and distracted users. This requires relatively sophisticated AI/ML that can track, follow, and navigate a chaotic environment where unintended collisions may be unavoidable.

Small Consumer AMR AI/ML Vision Systems & Proximity Sensors

While a 2D camera image can provide context, a proximity system is still needed to determine distance and depth from physical objects. Therefore, sensor fusion is critical for small AMR luggage, which is expected to zip alongside a user without excessive guidance.

Sensor fusion is a common method for a robotic system to gain greater insight into the environment than a regular 2D camera image would reveal. Robotic vision systems used to be rather bulky and expensive but have progressively shrunk as vision systems, and proximity sensor packages have become more compact and more readily used for machine vision.

For AMR luggage to maintain the svelte aesthetic appeal of modern luggage, the proximity sensors and vision system must be compact enough to fit within a smooth profile without being too overt. This puts additional space constraints on the vision and proximity systems and necessitates greater levels of integration, possibly with a collective vision and proximity system on a single compact PCB. This approach toward hyper-compactness may also be necessary to reduce the weight associated with AMR electronics for manual use.

Hence, a time-of-flight sensor that uses invisible 940nm infrared (IR) light in a small package could be ideal for such an application. IR sensors can be protected by optically clear lenses at IR frequencies but relatively opaque at visual light frequencies. Moreover, target color and reflectance invariants include sensors that employ Vertical Cavity Surface Emitting Laser (VCSEL) IR technology for time-of-flight sensing (Figure 1).

Figure 1: Functional diagram of a VCSEL time-of-flight IR proximity sensor 511-VL53L5CXV0GC/1 ideal for use in AMR luggage applications. (Source: Mouser Electronics)

Small Consumer AMR Mobility

Mobility systems are a significant design focus for small AMRs. The responsiveness, efficiency, and reliability of a small AMR system are largely dictated by the quality of the design and choice of devices/components that go into the AMR mobility system. A typical AMR mobility system includes electric motors, motor drivers, and motor control system, as well as the energy storage, usually an electric battery. Highly efficient motor control microcontroller units (MCUs) are likely the best option for such a compact and low-power application. Other motor control solutions may be less power efficient and have a larger footprint even if higher performance is possible, such as field-programmable gate array (FPGA) solutions.

A small consumer AMR luggage will likely use small, brushless DC (BLDC) motors, usually three phases, to achieve the responsiveness and efficiency desirable for this application. However, BLDC motors require a specialized motor controller and driver technology that can drive three separate phases simultaneously with sensor input capability and enough power to handle the computation load of BLDC motor control mathematics.

In the past, motor control systems like this required a complete custom design of each feature and the necessary resources to iteratively design a functional solution. Now, BLDC motor controllers with embedded MCUs provide a highly integrated solution that can easily be more compact and possibly even more efficient than prior less-integrated approaches (Figure 2). These highly integrated solutions aid development time, reduce PCB area, and minimize overall Bill of Materials (BOM) complexity.

Figure 2: A highly integrated motor controller/MCU, such as the STMicroelectronics STSPIN32G4, can provide advanced motor control features and processing in a single package. The image above is the STMicroelectronics EVSPIN32G4 Demonstration Board with the STSPIN32G4 system-in-package (SiP) and STL110N10F7 power MOSFETs. (Source: Mouser Electronics)

Small Consumer AMR Wireless Communication

Small consumer AMR luggage also needs to be readily provisioned and controlled by the user through ordinary means. Moreover, modern electronics can benefit from over-the-air (OTA) upgrades, especially with new and experimental products like AMR luggage. This means that AMR luggage systems also need to incorporate a variety of wireless standards to be compatible with a user's smartphone and possibly connect to other IEEE 802.15.4 wireless protocols for testing, diagnostics, or other smart home feature integration.

Given AMR luggage's space and power constraints, separate wireless chips can significantly increase board area, especially when considering the need to maintain electromagnetic compatibility (EMC) compliance with a design. Hence, a microcontroller with built-in wireless capability, such as Bluetooth^® Low Energy 5 (BLE5) and IEEE 802.15.4^™ communication protocols, could significantly ease the development process and enable more rapid development of a wireless capable AMR luggage (Figure 3).

Figure 3: A multi-protocol wireless microcontroller system-on-chip (SoC) can greatly benefit the design and development of IoT devices while extending application battery life. (Source: Mouser Electronics)

Conclusion

Though flying cars are still developing, robotic luggage that autonomously follows its user and eliminates some of the drudgeries of long-distance travel is here today. The future will see more innovative robotic luggage ideas emerge, which all must contend with fitting autonomous mobile robotics systems in a highly compact package while ensuring safety and user-friendliness.

An Airport Pioneers Fast Data Connections

Mon, 12 Sep 2022 14:51:11 GMT

(Source: icholakov - stock.adobe.com)

Nearly 15 million airline passengers per year pass through Dallas Love Field, one of Texas’s largest airports. Like most air travelers around the world, visitors to Love hope to make their connections—their data connections, that is.

Airports are often where people most need to communicate, whether to get an update on flight information, check in with the office, or reach out to family. But like many airports, Love hasn’t always been able to ensure high-speed Wi-Fi^® and cellular access for its visitors. During busy times, conventional Wi-Finetworks are slowed by thousands of people tying in at the same time and in the same area, many of them running high-bandwidth apps like music streaming, video meetings, and photo sharing. Meanwhile, airports typically have weak cellphone signals because phone carriers aren’t allowed to place cellphone towers anywhere near runways.

Experimenting with Spectrum

Airports aren’t the only places leaving users frustrated with slow or unreachable networks. Wi-Fi and phone carrier service are often weak or stretched thin at busy locations. “Whether in an airport, a stadium, a manufacturing facility, or an office building, most people want better connectivity,” says Derek Peterson, an engineer who serves as chief technology officer at the global network services provider Boingo.

Everyone complains about network speeds, but in 2018, Love did something about it. Working with Boingo, Love became the first major US airport to install a private LTE data network—essentially equivalent to running a private phone company in the airport. This network provides the same sort of high-speed data connections people are used to from their phone carriers, at least when they have a good signal.

For the airport’s network technology, Boingo and Love enlisted OnGo, an industry standard for the 3.5GHz Citizens Broadband Radio Service (CBRS) band, a currently underutilized piece of the radio spectrum. Using CBRS avoids having to try to license part of the spectrum used by phone carriers—an option far more costly and, in many areas, unavailable at any price. The OnGo network was so leading edge that Boingo and Love had to request a special temporary license from the Federal Communications Commission (FCC) to run it, with the understanding that the network was an experimental one to prove that it could operate without interfering with phone carrier and other signals. “We can adjust the signal power so it’s strong enough to blanket the location but not too strong that it causes problems for outside networks,” explains Peterson.

Just one hitch exists: Because CBRS is a new technology, most phones can’t access it right now. But 90 percent of smartphones will be ready to access CBRS by 2023. According to Peterson, CBRS networks are expected to become a major resource at venues worldwide over the next few years, placing Love far ahead of the game. Aside from helping to prove to the FCC that the networks can operate effectively without interfering with phone carriers, Love’s network has become an important resource for the airport’s internal operations. When the public’s phones are CBRS-equipped, Love’s network will be ready to serve them. Moreover, the network can easily be upgraded from LTE to the newer, higher-speed 5G standard rolling out around the world.

Another technology used for fast data transfer is silicon photonics. It is faster and can travel longer distances using both wave and particle behavior. The advancements in photonic integrated circuits are used to shrink the size of the circuit board by eliminating the parallel buses for optical links. These higher speeds over long distances with the smaller chip size have been made possible by Intel’s hybrid silicon laser. The higher power, more costly dedicated transceiver modules increase bandwidth at higher density with less power by directly converting the electrical to light. Intel^® Silicon Photonics 800G DR8 OSFP (Optical Small Form Pluggable) Optical Transceiver modules can overcome the volume shortage of conventional short and long-haul optics modules. Volume product is available today at 100, 200, 400, and 800GB/sec.

On-Demand Capacity

In the meantime, Love is taking other measures to ensure high-speed data access at the airport. The private LTE network is just one component of a three-pronged initiative to solve the connection problem. Love also boosted cellphone service for visitors by setting up a distributed antenna system (DAS)—a series of mini cellphone towers scattered throughout the airport to pass on stronger signals from major carriers. And Love upgraded its Wi-Fi network to Passpoint^®, a Wi-Fi standard allowing phones to jump onto the network without any sign-on.

Love’s efforts provide an advanced look at where data access is heading everywhere, says Peterson. That future, he explains, is one where access to high-speed connections is much more fluid and efficient, with phones constantly and seamlessly switching in the blink of an eye between different cellphone bands, private phone networks, and Wi-Fi networks to hunt down the best connection.

At the same time, he adds, the cellphone system won’t limit a carrier to a specific piece of the spectrum or to particular cellphone towers and equipment in the future. Instead, he says, spectrum and equipment will be dynamically allocated moment to moment to whichever network needs the extra capacity. “You don’t always have to build new networks,” says Peterson. “The existing networks can communicate with each other and with phones to create the right connection for the traffic at a given location and time.”

These changes should make all airports—along with other busy, data-starved venues—places where people can get the fast connections they have come to depend on. Your flight might still be late, but at least you’ll be able to let others know and watch a movie to pass the time.

Interior Vehicle Sensing AI Improves Safety

Fri, 01 Jul 2022 05:01:00 GMT

(Source: Mike Mareen - stock.adobe.com)

Today we hear about self-driving vehicles on the horizon, but true autonomous driving in varied real-world conditions is still many years away. Human drivers still need to be attentive to the situation at hand, and the interior of a vehicle appears to present a relatively static lab-like environment for observation. Eyeris, started in 2013 as a human-centric artificial intelligence (AI) company, aims to make driving safer and more comfortable by monitoring interior conditions, ensuring that the human is in control, and confirming that even the environment is up to performing this critical task.

The Challenge: Varied Occupants and Sensing Conditions

While relatively static compared to the outside world, vehicle interior conditions still present a variety of challenges. A person might be driving the car alone, or several additional occupants could be in the car who might be male or female and range in size from small children to 100-kilogram adults and beyond. Add on to this the fact that humans have a wide range of skin tones and might be wearing different clothing and accessories in different lighting conditions and temperatures, and suddenly this “lab environment” becomes a rather complicated experiment. That is even before considering a family pet or two along for the ride, the hamburger wrapper in the back seat that wasn’t cleaned up yesterday, and a phone or two dropped in the passenger seat.

The Solution: Sensor Fusion and Data Abundance

While one sensor system might boast the best eye-tracking or other technical merits, as an AI software company Eyeris instead focuses on fusing a variety of hardware sensing elements. As such, they partner with a wide range of hardware manufacturers for sensing technologies—including traditional infrared (IR) modern red, green, blue, plus infrared (RGBIR) sensors, thermal imagers, and even radar—to get an overall view of the situation and collaborate with a wide range of processor manufacturers to run AI routines. This sensor fusion, combined with an extremely large dataset used for training, means that the interior space of a vehicle can be accurately interpreted in the same way a human amalgamates sight, hearing, touch, smell, and possibly even taste to perform a complicated task.

In addition to the raw computing power needed to run an AI system, connections among camera hardware, sensor processing modules, and the other processing hardware of an automobile also must be considered. For instance, Eyeris has used for some of its reference designs a Maxim’s MAX96706 deserializer to connect mobile industry processor interface (MIPI)-based image sensors and camera modules into the AI processing board with great success. As automotive electronics become ever more integrated, reliable methods of handling and abstracting this data transfer are important to consider.

The wide range of automobiles that are manufactured means that a well-organized system that can be easily integrated into automobiles X, Y, or Z can significantly reduce development costs and time to market.

Hardware Innovation: Facilitating Software Innovation

We have seen an incredible explosion of computing power and hardware innovation over the past decades. That being said, innovation cycles for software naturally move at a much faster pace than those for hardware, and manufacturers often find themselves in a “catch-up” mode in relation to their software counterparts. It is one reason Tesla, Apple, and others make their own AI hardware to cater specifically to software improvements that are on the horizon.

For smaller software/AI companies, which partner with a wide range of existing hardware manufacturers, it is important to have mature software stacks and software development kits (SDKs) available that are compatible with the latest AI frameworks—such as TensorFlow, PyTorch, and ONNX—in addition to having adequate raw computing power. Available compilers should support modern neural network layers, with mature software emulators, simulation engines, and related tools for AI model parsing, pruning, quantization, and other tasks. Finally, enabling sensor fusion tasks, such as built-in 3D disparity engines, multi-camera streaming capabilities, rich input/output (IO) interfaces, and more are also incredibly helpful. This enables AI, and those that set up AI systems, to work with a broad array of data while cutting through the noise.

AI Sensor Fusion: Automotive Safety and More

While this blog focuses on interior automotive sensing, more generally speaking, a range of applications exist in which a traditional vision-only AI setup might seem like the logical choice but may not be sufficient for a particular use case. Especially in safety-critical applications, a vision system that works most of the time in proper lighting and other conditions may be far from sufficient. In these situations, adding additional sensing capabilities—whether that be a second RGB visible light device, an IR sensor, radar, or even something like a thermal sensor for enhanced presence detection—may enable AI to sufficiently monitor and control an environment.

Multibillion-dollar companies may have the resources to develop their own chips in-house, but in other situations, a smaller, more flexible AI company can be the right fit for the job. Here the proper hardware partners must be identified, developed, and integrated to produce an all-in-one product for automotive and other industries. The better the available hardware and software interfacing tools, the easier it is to set up AI software, and the faster an excellent product can be produced. With the proper data, tools, and AI training, we can make our world safer and better for the users of such systems and for society as a whole.

Securing the Edge in an Insecure World

Fri, 03 Jun 2022 14:35:04 GMT

(Source: Tierney - stock.adobe.com)

Of all industries and technologies, none have grown in scale as fast as the Internet of Things (IoT), and what started out as a handful of experimental internet-enabled microcontrollers in the early 2000s has grown to more than 21 billion devices worldwide. While Ethernet allowed devices to connect to the internet, Wi-Fi^® really fueled the IoT industry as it enabled very small devices to have internet connectivity no matter where they were.

The first devices were used for trivial applications such as wireless thermometers or humidity sensors—Interesting gizmos as opposed to practical products. With limited use and an extremely small market, virtually no security measures were used for these seemingly low-risk devices. But, as microcontrollers became more advanced, these simple IoT products started to do more and eventually could be used to stream video and audio. Even though these devices became increasingly more advanced, security was seriously lacking, with many IoT devices not using passwords, using insecure connections, and even storing private data unencrypted.

Fast forward to 2022: Billions of devices around the world are potentially vulnerable to cyberattacks, and hackers frequently target IoT devices for their private data and abilities to perform denial-of-service attacks. Protecting these devices with strong security practices has never been more important.

What Is Edge Computing?

As the first generation of IoT devices was based on extremely simple microcontrollers with limited capabilities, any heavy data processing had to be done remotely on servers (i.e., cloud computing). While this is advantageous from an energy perspective, it comes with some major challenges: Latency between submitting data and receiving a response, and the fact that potentially private data have to be streamed across the internet where they may be intercepted.

With edge computing, however, some or most of the heavy computation is done either on IoT devices or on a computer local to the IoT device. Such a system not only reduces the latency but also minimizes the risks associated with sending private data over remote networks. Edge computing can even be used to preprocess data (e.g., video camera feeds) so that any data that are sent to a remote connection will have private information obscured or removed.

How Are Devices Attacked?

Even with edge computing, IoT devices are still vulnerable to a wide range of attacks, and engineers must understand these attack methods in order to defend against them.

Data Vulnerabilities

The most important concept to be understood is the three states of data: In storage, in transit, and in processing. Data are vulnerable in all three of these states, but data are particularly vulnerable during transitions between these states. For example, data stored in memory can be accessed, data being transferred between the central processing unit (CPU) and memory can be peeked, and data inside a CPU may be accessible via side-channel attacks.

Poor Protocol Implementation

Poor protocol implementation is often a cause for concern. For example, although the theory behind OpenSSL is perfectly adequate for protecting data, its implementation revealed a major bug called Heartbleed that allowed for buffer overflow attacks to return private data in a server’s memory.

Exposed Programming Ports

Many devices on the market require a programming port for flashing program read-only memory during manufacture, but hackers can access this port to gain access to the main microcontroller (Figure 1). From there, program memory can be dumped, IP stolen, and keys obtained.

Poor Use of Operating Systems

Some devices use complex security operations centers capable of running operating systems such as Linux. While this allows designers to create complex applications, these operating systems may have exposed ports that are not closed by default, use well-known root passwords, integrate software packages that are not needed but contain bugs, and are not kept up to date.

Lack of Device Authentication

IoT designs that need to use a remote server for data storage and communication rarely contain a certificate system to authenticate the IoT device. Any device pretending to be an authentic IoT device can potentially access an IoT server. Even worse, a device could be cloned with inbuilt malware and infect a remote server.

Figure 1: Microcontrollers often have easily accessed pins and programming ports. (Source: zdyma4 - stock.adobe.com)

How Can Devices Be Protected?

Unfortunately, with so many attack vectors, no catch-all solution to protect devices exists. However, common sense and the use of hardware security measures can dramatically improve the security of a device.

The first line of defense is to use encryption whenever possible, as encrypted data cannot be read without the key. Encryption can be applied to memory, to data in transit between system components, and to data being sent across the internet. The second line of defense is to use strong security algorithms and routines (e.g., true random number generators and encryption algorithms) that do not have poor implementations.

Of course, these two methods of defense require that encryption keys are kept safe from hackers and that algorithms are immutable. Fortunately, hardware security coprocessors exist for this very purpose, and a good example of such a device is the EdgeLock SE050 by NXP. This coprocessor integrates true random number generation, relies on multiple encryption algorithms including AES, RSA, and DES, can be used to store keys, and supports trusted platform module capabilities (Figure 2).

Figure 2: The EdgeLock SE050 offers multi-layered protection against attack. (Source: NXP Datasheet)

Conclusion

As cyberattacks on IoT devices continue to increase, the need to secure these devices with strong security practices has never been more important. Software security measures can only go so far and trying to use software implementations for encryption can lead to disaster if not kept updated. Hardware security solutions such as the NXP EdgeLock SE050 are ideal for securing devices.

Depth Sensors Visualize Volumes

Fri, 15 Apr 2022 05:01:00 GMT

(Source: HQUALITY - stock.adobe.com)

Robots and machine vision devices use a variety of feedback mechanisms to assure accuracy. Discerning 3-D spaces with reasonably good accuracy can be done in several ways. So far, optical, sonic, and mechanical vision and sensing techniques have been employed with adequate levels of success. More demanding requirements force us to sharpen our already sharp pencils. Next-generation vision-based designs center on more precise depth and volume sensing with higher accuracy.

Common Sense

Techniques used so far have been okay for the problems at hand. The lowest cost mechanical depth or surface sensors can be as simple as a spring-loaded linear trimpot or limit switch.

When it comes to precision, sonic and optical techniques have proven more resolute without the need for moving parts. Optical distance sensing is used for simple proximity detection and more precision range finding. Proximity from less than a millimeter to 8 meters can be discerned as a digital go/no-go signal indicating the presence or absence of a target.

Thanks to the low-cost, high-resolution modern generation of cameras, video techniques have been at the forefront of distance and volumetric measurement. Next-generation design and requirements push device manufacturers to offer high-performance solutions.

More Need for More Options

In addition to machines needing higher precisions, the post-pandemic world created the need to detect people and occupancy numbers in a given location. Spacing between people is a relatively new requirement many must incorporate. A similar condition requiring attention is that of dementia. In assisted living facilities, the ability for an all-encompassing computer system to track wandering patients’ locations is crucial.

Industrial and factory applications too can take advantage of more robust and accurate distance and volume measurement subsystems. As more advanced fabrication technologies progress and merge, feedback on accuracy, position, direction, speed, and depth becomes critical for next-generation fabrication machines. For example, milling machines rely on precise motors and gear assemblies to correctly position cutting and grinding heads. Too deep, and a cutting head will break. Too shallow, and too much material is left. These machines will hit the right spot with accurate distance sensing, even if calibration is off. Closed-loop feedback produces better results. CNC machines, 3-D printers, and laser/plasma cutting and welding machines also benefit from higher accuracies of closed-loop feedback.

Newest Innovations

Analog Devices anticipated a growing need for volumetric sensing and measurement over many applications. The AD-FXTOF1-EBZ is a dedicated modular video engine with embedded Time of Flight (ToF) distance measurement built-in (Figure 1).

Figure 1: The modular 3-D sensing development kit supports various applications from volumetric measurement to occupancy and activity detection. (Source: Analog Devices)

The VGA resolution of 640x480 at 30 frames per second allows easy integration as a peripheral function to a host application supervisor. It features a two-lane Mobile Industry Processor Interface (MIPI) that can use a 25-pin or 15-pin flex cable to an interposer board.

The 940nm IR laser is an eye-safe vertical-cavity surface-emitting laser (VCSEL), which reduces manufacturing costs by eliminating the right-angle emitter configuration. It also touts its ability to operate in high light conditions thanks in part to the optical 940nm bandpass filter. This helps block noise and interference from external sources. A batwing style diffuser is used to provide the receiving lens with a precise 87-degree by 67-degree field of view.

Performance-wise, the video depth finder has two settable ranges it can operate within. A 20cm to 180cm range and a 50cm to 300cm range maintain a 2 percent accuracy. It will require a 5V 2A power supply rated from -20ºC to +75ºC, making it somewhat tough and rugged environmentally.

The SDK development kit style interface allows it to connect to a host microprocessor, microcontroller, or single-board computers like Raspberry Pi or Nvidia (Figure 2). The SDK also provides OpenCV, Open C/C++, Python^®, MATLAB^®, Open3D, and RoS wrappers so that developers can use them to simplify application development. Connection options include USB, Ethernet, or Wi-Fi, and reference design and bill of materials are available.

Figure 2: The actual camera and lens RFPC board and image processing AFE boards use IIC interfaces for control and configuration. Operational GPIO and MIPI interfaces allow real-time control and data access. (Source: Analog Devices)

Conclusion

The need for fast and resolute image and distance sensing allows designers to create next-generation sensors, robots, vehicles, and safety systems. The AD-FXTOF1-EBZ from Analog Devices lets you test the water quickly and easily. Expect higher resolution, faster frame rates, and longer distances with future versions of this technology as it gets adopted across different home and industrial applications.

Gateway Implementations with RISC-V

Fri, 04 Mar 2022 20:38:49 GMT

(Source: a-image/Shutterstock.com)

Reduced instruction-set architectures (ISAs) such as RISC-V provide greater efficiency and less drag on resources than their more complex counterparts. Industrial Internet of Things (IIoT) applications often require both high connectivity and cooperation levels between modules while keeping costs down and reducing power consumption. The Terasic T-Core FPGA MAX 10 Development Board provides a comprehensive hardware design platform built around the Intel^® MAX 10 FPGA for RISC-V-based designs. It’s an optimal development solution for cost-effective designs in control plane or data path applications and features industry-leading programmable logic for design flexibility.

Gateways in IIoT Applications

An Internet of Things (IoT) gateway combines and bridges a variety of sensor readings—often using analog, digital, or simple serial communications—to higher-level serial communications channels such as a simple UART, more complex channels like I²C or SSI, or even CAN, USB, or Ethernet. This bridge often does some local computation so that raw data doesn’t need to be sent to the cloud—instead, a notification is sent when a sensor reading moves out of range.

A development platform for such an IoT bridge requires a significant amount of flexibility—on the sensor side supporting a variety of analog inputs, general purpose inputs, and simple serial communications; and on the management side providing higher level communications (such as I²C, and SSI)—while providing computational and storage capability for data processing.

An ideal target development board for this type of bridge is the Terasic Technologies T-Core FPGA MAX 10 Development Board (Figure 1). The MAX 10 FPGA can implement many standard serial interfaces programmable logic elements. The FPGA can also host a RISC-V core for processing, and the board has an off board QSPI flash device for source code and data storage. The FPGA has dual ADCs, with up to 10 pins for sensor readings. The board has 12 I/O pins for either general purpose use or use as I²C or SSI communications channels.

Figure 1: The T-Core FPGA MAX 10 Development Board (Source: Mouser Electronics)

Implementing RISC-V for Bridging Applications on the Terasic T-Core FPGA MAX 10 Development Board

Implementing the efficient RISC-V processor on the development board directly aligns with many of the key requirements of an IoT bridge. The most critical aspects include increased efficiency in power and processing, lower costs, wide protocol flexibility, and strong security.

Efficiency

One of the fundamental advantages of the RISC-V ISA is its processing efficiency. Simple CPU operations use memory directly without specialized processor registers, increasing speed and reducing the required memory footprint. With a cache subsystem, frequently used locations are automatically available with reduced access times—reaping the benefits of fast specialized register access without complicated and less efficient coding. Gateways benefit from this advantage with low power and small code space. Also, gateways are very data-transfer intensive because data packets are typically only transferred, broken down, or stitched together. Minimal processing is needed to change from one protocol to another, making efficient memory movement a key benefit. More efficient processing also helps implement AI-oriented gateway functions to identify unusual events and predict potential issues before they become problems.

Flexibility and Protocol Support

Gateways need to be flexible at the protocol, operating system, and in physical connectivity, and modular in construction. The RISC-V open-source architecture makes it easy to support various protocols and adapt to changing requirements. Accessing the source code for peripheral drivers and stacks and the associated protocols makes it easy to modify them as needed, both during development and even after deployment. This makes it easy to modularize peripherals and protocols so they can be easily swapped, updated, or enhanced as industry standards change. This can extend an IIoT gateway’s lifetime and reduce the overall system deployment cost—a key factor in IIoT implementations.

Security

RISC-V hardware-based security is needed to implement the root of trust, the bedrock of any robust security system. The root of trust is the known secure starting point for a host of security-related functions such as secure boot, cryptographic computations, secure key, and certificate storage. The root of trust is commonly supported with specialized hardware for protecting secured data and peripheral functions, implementing tamper protection, generating keys, and providing secure updates to application software. When a system requires cloud storage, the gateway can use trusted cryptographic standards to protect data to and from the cloud (Figure 2). With open-source implementations available for encryption, decryption, certificate management, and secure data communication protocols, the developer has access to all the security-related code, making it easier to test and verify the design’s robustness. Additionally, the ability to customize and upgrade the code as needed for specific application requirements—without the need to wait for a third party to develop and release periodic updates—is an additional benefit of an open-source environment.

Figure 2: The gateway can use trusted cryptographic standards to protect data to and from the cloud. (Source: sdecoret/Shutterstock.com)

Conclusion

Gateways will continue to evolve as the IIoT environment produces new applications and revenue streams. As they change and become more complex, additional processing power will be required, meaning more data processing within the gateway to minimize data traffic to the cloud will also be needed. The Terasic T-Core FPGA MAX 10 Development Board can provide developers with the tools they’ll need to design cost-effective, single-chip solutions for these data-intensive applications. The out-of-the-box RISC-V support available with the kit is conducive with the efficiency, flexibility, and security required for IoT bridges in the present and future.