Accelerating the Development of “Smart” Industrial Equipment
The global landscape of smart, IoT devices is on the brink of a monumental shift. By 2033, their numbers are expected to nearly double, jumping from 15.9 billion in 2023 to over 32.1 billion. This surge is fueled by a series of technological advancements. Innovations in processing, sensing, and wireless communications are at the forefront. Meanwhile, strides in data analytics, particularly in artificial intelligence (AI) and Machine Learning (ML), are playing a crucial role in this expansion.
Time to revenue is critical — but developers face several challenges while integrating these multiple technologies.
This article elaborates these challenges before discussing how the Arduino product portfolio and ecosystem accelerates time to market, finishing with a case study where the Arduino ecosystem enabled the development of a smart helmet for military and industrial use.
Figure 1: Powerful multi-core microprocessors are available in tiny form factors.
Enabling technologies
A range of core technologies are enabling growth in smart devices. Powerful microprocessors are available in tiny form factors, with multi-core ARM® Cortex® - based processors (Figure 1) traditionally favoured by embedded device developers.
The software deployed in IoT applications has also become more sophisticated, enabled by the embedded processing power. Integration with cloud computing platforms, such as Amazon Web Services (AWS) and Microsoft Azure, has enabled organisations to build, scale, and secure IoT solutions, while capturing and storing the vast amounts of data generated by the smart devices ‘at the edge’.
Big data analytics techniques, such as Artificial Intelligence (AI), have emerged to leverage this stored data and Machine Learning (ML) algorithms deployed at the edge act on sensor data in real-time.
Developing a smart device therefore requires skills and knowledge across a wide range of technologies, presenting several challenges to the developer.
Developing smart devices
For the developer who wants to focus on the requirements of the application and get to market quickly, the prospect of acquiring or assembling all necessary skills and expertise can be daunting. Skills, such as security expertise and AI programming are in short supply, and the cost and time of hiring may kill a project.
At some stage the developer must decide between a ‘make’ or ‘buy approach but the answer may not be evident up front. Pre-integrated hardware reduces non-recurring engineering costs (NRE), but bill-of-material (BoM) costs will rise quickly with volume. A ‘chip-down’ approach will increase upfront NRE costs and extend development cycles, but, at a certain production volume, BoM costs will be lower and NRE can be amortised over a larger volume of devices.
Fortunately, an ecosystem has developed within the IoT supply chain which addresses many of these challenges. Semiconductor manufacturers and third parties, such as Arduino, have developed products that integrate much of the required functionality, accelerating evaluation, prototyping, and even medium volume production runs. System-on-Modules, (SOM), such as the Arduino’s Portenta H7, integrate processing capability with functionality such as DRAM, boot-flash, voltage distribution, and communications interfaces.
Software development can also be accelerated through use of open-source libraries allowing capable developers to adapt existing software for their own needs, while providers such as AWS and Microsoft offer a range of tools to simplify the development of cloud-based solutions, including AI applications. Tools such as Edge Impulse facilitate the development of ML algorithms which access and act on sensor data.
In the current IoT environment, where speed-to-market is critical, an effective ecosystem enables the solution developer to reduce the costs and timescales of the development cycle by leveraging the products, solutions, and skills of their ecosystem partners. An ecosystem partner like Arduino can offer support in numerous areas, including hardware, software tools and libraries, reference designs and tutorials, and much more.
Figure 2: The Arduino Pro ecosystem.
Arduino Pro product set and ecosystem
With the launch in 2020 of the Arduino Pro product range, Arduino cemented its arrival as a member of the industrial IoT ecosystem. The Arduino Pro family consists of a range of industrial grade SOMs, smart sensor boards, industrial automation controllers, and connectivity boards supporting the development of edge computing, AI applications.
The range, which includes the Portenta SOM series, and the Nicla Sense ME and Nicla Vision boards, complements Arduino’s extensive Interactive Development Environment (IDE) with the comprehensive Arduino ecosystem — also including a range of pro development kits and cloud connection tools (Figure 2).
The Portenta series of high-performance, industry rated SOM boards integrate dual Cortex® ARM® microcontrollers with Wi-Fi and Bluetooth® connectivity and other peripheral functionality, such as the powerful Chrom-ART graphical hardware acceleratorTM on the Portenta H7. The boards support high-level language programming and AI, while performing low-latency operations on customisable hardware, and offer a range of vendor options for the main processor, including STM Microelectronics, NXP, and Renesas.
The Nicla Sense ME combines a mix of sensors, including an integrated motion sensor, magnetometer, pressure sensor, and unique 4-in-1 gas sensor with integrated high-linearity, and high-accuracy pressure, humidity, and temperature sensors. This tiny board, developed with Bosch® Sensortec, features a 9DoF smart motion sensor and a 4DoF environmental sensor with AI capabilities.
The Nicla Vision sensor board features an ultra-compact 2 MP colour camera with the intelligence to process and extract useful information from anything it sees. The board also includes a smart 6-axis motion sensor, integrated microphone, and distance sensor, enabling capture of distance, sound, movement, and vibration data — this makes it ideally suited to asset tracking, image detection, object recognition, and predictive maintenance applications. With a powerful STM32H747AII6 Dual Arm® Cortex® processor, the Nicla Vision keeps energy consumption low and can even be powered by battery for standalone applications.
The Arduino Pro products complement the extensive Arduino open-source platform that has supported thousands of projects over the years, ranging from everyday objects to complex scientific instruments. Over 50 million Arduino products have been sold since the company was founded and, out of over 83 million projects deployed globally, over 2,000 have been initiated by enterprise customers. With over 30 million active community users, 500,000 LinkedIn users mention Arduino as a core skill and 2 out of 5 engineers have Arduino experience.
Developers of industrial IoT applications are now able to access this rich ecosystem, not just the products but also the knowledge embedded in the Arduino IDE, the reference designs and tutorials in the range of Pro Kits, and, crucially, the skill sets available on the market. OEMs leveraging Arduino’s open-source platform can accelerate time to revenue by up to a factor of 5 (Figure 3).
At the same time, Arduino’s open architecture includes a powerful middleware platform that enables a level of abstraction from the underlying hardware. The developer can focus on creating the application code, without becoming involved in underlying hardware interfaces and drivers, OS and security APIs, and so on.
Also, if the underlying SOM board needs to be changed — for example, from one with an NXP processor to a Renesas-based board — the process of adapting the code is hugely simplified, taking hours rather than days or weeks. The middleware platform therefore accelerates product development while also supporting hardware portability, reducing product risk by bringing a level of future-proofing and preventing vendor lock-in.
Figure 3: The Arduino ecosystem can accelerate time to revenue by as much as a factor of 5.
Larger volume production runs may mean that a chip-down approach becomes more effective at some point during the lifecycle of the product. In this case, where a product launch is initially based on the Arduino ecosystem, Arduino can work with the customer to depopulate the design to support chip-down manufacturing.
Smart Helmet case study
Numorpho Cybernetic Systems, a Chicago-based innovator, were recently commissioned by the US DoD and a smart manufacturing association to design an ecosystem of smart products, including a smart helmet. The helmet was intended for applications in military, construction, healthcare, and consumer sectors, its key requirements being to reduce operational costs related to workforce safety by providing environmental data that could be processed in real-time, at the edge. The helmet had to meet stringent government and environmental regulations, and Numorpho faced contractual penalties and reputational damage if they failed to deliver against ambitious project timeframes.
Numorpho approached Arduino who identified the key criteria for a solution design as:
Figure 4: The solution was based on the NiclaSense ME and Nicla Vision sensor boards along with the Portenta H7 SOM.
- Durability and adaptability for a wide range of industrial settings
- High-reliability and accurate monitoring of local environment to provide real-time awareness and proactive risk management
- Seamless system integration and flexibility for future upgrade
- Three-month development cycle to produce prototype
The solution agreed with Numorpho was based on the NiclaSense ME and Nicla Vision sensor boards along with the Portenta H7 SOM (Figure 4). Development times were accelerated by leveraging the Arduino IDE and Cloud Connectivity achieved using the Arduino Cloud solution. The tiny form factors of the Arduino Pro hardware simplified physical integration within the helmet and the communications interfaces enabled connectivity to a Numorpho Mobile App while enabling Edge Impulse to inject ML routines as the helmet learned about its environment.
The Arduino solution enabled Numorpho to successfully meet their client’s exacting requirements, avoiding the need for a chip-down design saving an estimated $110,000 in NRE costs and 20% in R&D costs. The solution proved Arduino Pro’s industry grade hardware, unlocking the Arduino ecosystem for industrial use cases.
Speed to revenue requires a strong ecosystem
Explosive growth offers boundless opportunity to manufacturers of smart devices, but the breadth of skill sets and knowledge required can add cost and time to the development process. The Arduino ecosystem enables the solution developers to leverage the products, solutions, and the open-source architecture and IDE to accelerate time to revenue by as much as 5 times, when compared with a chip down approach.
See Arduino Overview
About Author
Harvey Wilson
Harvey Wilson is a Systems Engineer Professional (Smart Industry) for Avnet Silica in the EMEA regio...
silica content library/articles/accelerating-the-development-of-smart-industrial-equipment
Accelerating the Development of “Smart” Industrial Equipment | Avnet Silica
