embedded award 2023: Hardware nominees
Increasing softwareisation does not change the fact that hardware platforms are the basis for every embedded system. The fact that this category had the most entries for the embedded award this year underlines the importance of embedded hardware products. These three nominees convinced our jury.
Integration of edge and cloud, the world’s smallest PM2.5 air quality sensor and a new type of audio processor
AWS IoT SiteWise Edge integration with CloudRail
Manufacturing companies use Industrial Internet of Things (IIoT) applications, such as Condition Monitoring or Predictive Maintenance to gain insights, optimize production processes, or reduce waste. However, connecting factory equipment and machinery to the cloud is a challenging task. This is especially true for older machines in the factory that lack modern communication standards.
Moreover, industrial customers need access to raw and high-fidelity sensor data to monitor their equipment in near real time. Often the required data just doesn’t exist with the existing sensors on the equipment or is not accessible. As a result, the majority of IIoT projects tend to fail because of the high effort and investment cost required to collect necessary machine data.
The combination of CloudRail with AWS IoT SiteWise Edge provides support for over 12,000 industrial sensor models, which can be used to easily gather data like vibration, temperature, pressure, velocity and much more.
With AWS IoT SiteWise Edge and CloudRail, industrial customers can now connect their assets across entire factories to the AWS Cloud using a fast and simple Plug-and-Play approach.
This integration provides a scalable and repeatable approach to process, filter, transform, and visualize that data at the edge. The solution comes with a fully managed gateway device, which unlocks existing data from sources like PLCs or factory management systems using a wide range of protocols and drivers such as OPC-UA, Siemens S7 or Modbus RTU.
Industrial assets, where no data is available yet or inaccessible due to vendor or security restrictions, are retrofitted with secondary sensors. These secondary sensors can be added to machines in order to collect data without requiring access to or interfering with the customers control systems.
BMV080 – Particulate Matter Sensor
Exhibitor: Bosch Sensortec
Clean air is fundamental to our health and well-being. Today, people typically spend approximately 90% of their time indoors, which has a negative long-term impact on health since indoor air can be three to five times more polluted than in the outdoor environment in most parts of the world.
A major component of this pollution consists of fine dust known as PM2.5 particulates, which are solid and liquid particles smaller than 2.5 µm. These particulates are inhaled and enter the lungs, where they can cause serious health problems.
The most common air quality measurement stations only provide outdoor data, which fails to deliver real-time, localized information most relevant to our health. Bosch Sensortec is addressing this issue with its new BMV080, the world’s smallest PM2.5 air quality sensor, which provides accurate, actionable data.
The innovation brings awareness and choice to people who value healthy living. The new sensor is easily integrated into ultra-compact IoT-based smart home and smart building systems to deliver air quality notifications to residents when PM2.5 levels rise above a threshold level. To take action, the resident or home automation system can ventilate, switch on an exhaust fan or filter to clear the air.
The BMV080 sensing element measures merely 4.2 x 3.5 x 3 mm³ (W x L x H), which is more than 450 times smaller than any comparable device on the market. The miniscule size and fanless, non-intrusive design of the BMV080 are a game-changer, and its noiseless operation means it does not disturb residents who may be sleeping or working.
The new sensor is maintenance-free with no fan-induced dust buildup, which means that it’s more reliable, and less prone to malfunction. It is a natural fit for ultra-compact wearable or IoT devices such as air quality monitors, smart thermostats, smart speakers, smart switches and smart air purifiers.
Comparable state-of-the-art PM2.5 sensors all rely on a fan to draw air across a laser beam, where a photodiode then detects and measures the particulates as they reflect or block the light from the laser.
The BMV080 utilizes an innovative design based on ultra-compact lasers with integrated photodiodes. It applies sophisticated algorithms to measure the PM2.5 concentration directly in free space, without requiring a fan.
Exhibitor: GreenWaves Technologies
The next generation of hearable products will achieve a new level of audio quality, adapting themselves to their wearer and their wearer’s environment. This is making new demands on embedded audio processors: a combination of DSP and AI workloads, executed at ultra low energy and latency levels yet retaining flexibility and ease of programming. At the same time, IoT sensors are requiring extreme processing capabilities while retaining a battery life measured in years.
GreenWaves’ GAP9 is a completely new type of processor designed from the ground up to address these requirements. GAP9 is an ultra-low power and easy-to-program processor that is able to process classic and ultra-low latency Digital Signal Processing (DSP) and Neural Network workloads in highly energy-constrained devices.
GAP processors are being used in battery powered IoT products that analyze images, sounds and more such as the kontact.io Portal Beam and hearable products that combine features such as AI-adapted Active Noise Cancellation, AI-based noise reduction and spatial audio.
Recent MLCommons MLPerf Tiny benchmarks results showed that GAP9 delivers industry-leading performance on a range of neural network tasks achieving the lowest latency and maximum energy performance of all entrants.
GAP9 delivers the best energy performance available on the market today for blended DSP and AI workloads without recourse to complex, opaque black box solutions that are a drag on product development time and cost. GAP9 integrates a 9 core, general purpose compute cluster with a highly integrated processing block for neural network processing (NE16).
This tight coupling of a flexible, programmable compute cluster and NN accelerator with shared memory and a common event system provides a level of flexibility unique in the market. GAP9 delivers best-in-class energy efficiency: not only on NN benchmarks but at the system level in real life, can uniquely run mixed precision (from 16 bit Floating Point to sub byte fixed point) NN, can run multiple NN at the same time, can handle heavy-duty DSP tasks such as LC3 codecs.
GAP9 also incorporates a highly flexible, sample by sample, ultra low latency audio filtering capability through GreenWaves’ Smart Filtering Unit. The coefficients in graphs running in this block can be dynamically updated providing a unique fusion of low latency filtering with AI parameter update enabling the next generation of ANC and audio transparency features.
GAP9’s use of the flexible RISC-V ISA along with its very homogeneous architecture allows us to provide a modern development environment that greatly improves programmer productivity. As one customer has told us, ”We do in weeks with GAP9 what we do with other platforms in months.