IoT Security Concerns Push Vendors to the Edge

　　Doing more processing at the edge to avoid sending sensitive data to the cloud emerged as a common theme among vendors at the Embedded World conference here last week. Whether this is a result of forthcoming GDPR (General Data Protection Regulation) laws coming into force across the European Union on May 25, or whether there it is simply a lack of sufficient security in current devices is difficult to tell.

　　A recent report on IoT cybersecurity readiness certainly points to the latter. Lawrence Munro, vice president SpiderLabs at Trustwave, who released the  IoT Cybersecurity Readiness Report, said, “As IoT adoption continues to proliferate, manufactures of IoT are sidestepping security fundamentals as they rush to bring products to market. We are seeing lack of familiarity with secure coding concepts resulting in vulnerabilities, some of them a decade old, incorporated into final designs. Because updating IoT devices by nature is more challenging, many remain vulnerable even after patches are issued, and often patches are not even developed. Organizations need to properly document and test each internet-connected device on their network or face introducing potentially thousands of new attack vectors easily exploitable by cybercriminals."

　　He added, “"Any device or sensor with an IP address connected to a corporate network may open the doors to a devastating security incident."

　　The report finds that while IoT use is growing rapidly — by the end of 2018, five in six organizations will be using at least a minimal level of IoT technology — security concerns are cited as the top barrier to increased IoT adoption.

　　Given this, it’s not surprising that there were several announcements and demos around IoT edge computing and security at Embedded World. On top of this, major vendors were also keen to announce their full IoT ecosystems or platforms, presumably to contain the security risks in a closed loop environment.

　　IoT on a Chip, with Multiprotocol Radio

　　As we highlighted previously, NXP was keen to illustrate the need for IoT edge compute capability.

　　“Self-learning edge node systems are going to start influencing IoT, with scalable processing at the gateway and the edge," said Geoff Lees, senior vice president and general manager of microcontrollers at NXP, during the company's press conference at the show. "The edge computing future demands high-performance compute together with secure data collection, management, and decision-making.”

　　Among the devices NXP showed were its ultra-compact IoT-on-a-chip package, which brings together the pre-integrated features from NXP’s i.MX applications processor family, and Wi-Fi/Bluetooth solutions for consumer and industrial developers to quickly build compact IoT products. Lees said the company’s proprietary ultra-low leakage SRAM with 10x lower static leakage than conventional memories, working with Samsung's non-volatile STT-MRAM that has 1,000-times faster wake-up time, 400 times lower write power than conventional embedded flash and 100 times longer battery lifetime, makes "instant-on" IoT edge node products possible.

　　STMicroelectronics demonstrated its new STM32WB wireless SoC, which combines microcontroller and the ability to run concurrent multiprotocol standards for IoT. The devices combine a fully-featured Arm Cortex?-M4-based microcontroller to run the main application as well as an Arm Cortex-M0+ core to offload the main processor and offer real time operation on the Bluetooth Low Energy (BLE) 5 and IEEE 802.15.4 radio. The radio can also run other wireless protocols concurrently, including OpenThread, ZigBee, or proprietary protocols, giving even more options for connecting devices to IoT.

　　"The STM32WB series delivers the advanced integration and uncompromising dual-core performance that developers now need to meet relentless end-user demands for even better and more affordable smart connected objects," said Michel Buffa, general manager of ST's microcontroller division. “Moreover, compatibility with the STM32 development ecosystem brings design advantages that can significantly reduce time to market for new products like lights, fitness trackers, medical monitors, beacons, tags, security devices, and many others.”

　　Another emphasizing processing at the edge was GreenWaves Technologies (see AI Comes to Sensing Devices). Its GAP8 IoT application processor, built on the RISC-V based Parallel Ultra Low Power (PULP) computing open-source platform developed at the University of Bologna and ETH Zurich, enables development, deployment and autonomous operation of intelligent sensing devices that capture, analyze, classify and act on the fusion of rich data sources such as images, sounds or vibrations. GAP8 is optimized to execute a large spectrum of image and audio algorithms including convolutional neural network (CNN) inference, with extreme energy efficiency, thanks to an integrated 8-core computational cluster combined with a convolution hardware accelerator.

　　Dublin-based S3 Semiconductors, showed its SmartEdge platform for enabling cost-optimized, custom single-chip solutions that put performance at the edge of the IoT networks. The company said processing at the edge is a rapidly evolving trend that accelerates data analytics at the source and reduces cloud-based processing, delivering more responsive decision making.

　　Edge devices typically require sensitive sensor analog front ends coupled to data converters, microcontrollers running complex embedded software, with security and some form of wired or wireless communications interface. Bringing all these functions together, often with the need for local calibration and low-latency real-time control, represents a significant design and integration challenge. S3 says its platform provides a way of integrating sensing, calibrating, controlling and implementing communication functions into a single ASIC.

　　At its press conference, Kontron also emphasized the concerns of many in industry who don’t want to put their data in the cloud, which is why it is targeting on-premises computing and storage. It said that embedded servers, or fog servers (cloud-like systems but located on premises), will be a huge growth area. Kontron says it is enabling customers to put more intelligence at device level.

　　As part of this edge to fog strategy, Kontron introduced an embedded server, the ZINC CUBE C232, tailored for complex computing tasks requiring the processing and analyzing of huge amounts of data, such as machine learning or artificial intelligence. It was developed as a platform for industrial applications, analytics, artificial intelligence and embedded vision. The embedded server can be used for time sensitive networking with up to four ports, making it suitable for edge and fog computing. It is available with 7th Gen Intel server class processors and can be equipped with up to 32 GBytes ofDDR4 RAM.

　　Kontron also announced a collaboration with NXP to enable more intelligence and efficiency at the edge for next-generation industrial IoT implementations for cloud, edge computing and factory floor innovations — or what is known as Industry 4.0.

　　GE Power’s Automation and Controls business unveiled the next generation of industrial internet connected control (IICS) solutions at Embedded World in Nuremberg.

　　"We’re bringing intelligence as close to the machine as possible," said Vibhoosh Gupta, the company’s product management leader for the business. "The cloud might provide an elastic compute environment which is good for things like fleet management. However, for applications requiring real time analytics, the inherent latency resulting from sending data to the cloud means there is no choice but to move to the edge.”

　　GE's IICS solutions provide data storage, analysis and rapid conveyance from the edge of industrial systems. It says that only a small fraction of data gets used in industrial processes, so its system leverages GE's Predix analytics suite at the edge and in the cloud to better utilize process data to help reduce downtime, and lower operational costs.

　　Renesas Enters Ethernet Module Market

　　Renesas Electronics revealed at the conference its entry into the industrial Ethernet module market, which it said was a logical enhancement of its chip business. The company describes its new I-RJ45 as a complete solution for an industrial Ethernet interface, containing the entire electronics and software embedded in an RJ45 connector.

　　The solution incorporates a single- or dual-port RJ45 connector and supports various industrial network slave applications including sensors and transmitters, gateways, operator terminals and remote I/O. The solution comprises an intelligent RJ45 module that comes with qualified embedded software to support multiple industrial Ethernet protocol stacks. The software package and sample codes provide system manufacturers with a complete set of tools and frameworks to build their application without any additional up-front costs or complexity.

　　The I-RJ45 module solution is a ready-to-use hardware and software solution that includes a pre-certified PROFINET implementation and can be used for mass-produced applications. Renesas plans to support other industrial protocols including EtherCAT and EtherNet/IP in the future. With a general application programmable interface (API), the application can easily be connected to the protocol software and offers an integration path to other Renesas ASSP solutions.