How to Design a PCB for the Enterprise IoT
If you were asked what the basis of the enterprise internet of things (EIoT) is, what would you say? For those who are involved in IoT development, the answer would be embedded systems. Surely, these are embedded systems that make it possible to “squeeze out” the value from disorderly corporate data. This results in multiple companies preferring to customize their environment instead of buying off-the-shelf products. That’s why the share of IoT hardware significantly prevails over other product types in the IoT market.
It naturally unlocks new opportunities for PCB design, as well as new challenges. PCB design largely determines if this ecosystem meets business requirements for the IoT ecosystem, but it also affects the whole lifecycle of the future solution, determines its adaptability to changing business environments and keeps up with evolving trends. How do you build a reliable bridge between business goals and a board when designing enterprise IoT solutions? Read on to find out.
Five pillars of EIoT PCB design
Any IoT solution to be implemented in the enterprise is expected to take a competitive advantage from the data while having an adequate cost of ownership. It is extensively utilized to optimize business processes through data-driven decisions, extend the equipment lifecycle through predictive-maintenance capabilities and run advanced automation through real-time data processing.
Thus, any PCB design for an EIoT project is based on five pillars:
- Data acquisition from objects/the environment
- Data transfer to digital format and its processing
- Data-based intelligent reactions
- Deep data analysis
- Comprehensive connectivity
In parallel, a PCB designer must solve a number of issues primarily associated with the cost of the final solution, considering various operation scenarios as well as the working environment.
EIoT PCB design challenges and solutions
IoT ecosystems tend to become smarter, but it’s not the only challenge that PCB design has to solve. It also needs to address the ever-changing business environment and industry requests, which means setting priorities properly. Given this, we recommend approaching the implementation of PCB design from the perspective of the following high-level benefits.
Flexibility is one of the main vectors for the evolution of manufacturing today. The opportunity to adapt the solution to changing conditions largely determines its value on the market. This forces the EIoT solution to be multifunctional and offer multiple connectivity options. It has become valuable for embedded systems to include various interfaces, such as temperature, vibration and light sensors, audio and video I/O and a range of memory interfaces. To simplify the configuration and updates, it’s advisable to implement both USB and OTA. Such great flexibility allows application developers to bring up the required interfaces, seamlessly reorganize the operational process and experiment with the system when it already runs.
The same applies to connectivity. Multiple wireless protocols offer a range of connection and configuration options. However, they should be complementary. For instance, Zigbee allows for seamless automation in the manufacturing environment, while Wi-Fi promotes enhanced control. What’s essential here is to balance small distance and speed protocols with far distance and high speed to provide more options, like Bluetooth+ LTE-M.
Business requires more precise prediction through advanced analytics tools, leading the data to meet more requirements than ever before. Signal integrity must not be in doubt. The heterogeneous data may include analog signals, video streams and so on, which must be unified. The data from multiple protocols must be transformed into standard protocols like MQTT and OPC UA. For instance, the implementation of the ICP/CFX protocol helps with a uniform data format.
The next requirement is to provide an opportunity to connect edge devices to analyze and preprocess valuable data before sending it to the cloud. The value also reveals the reduction of the cost of connectivity usage by filtering the amount of raw data, as well as the reduction of latency through the delegation of some basic decision-making credentials to the edge devices.
In addition, more and more IoT applications implement neural networks for data analysis. This is a common way for IoT visual inspection through incredible accuracy and rapid ROI. The snag is that a neural network is a resource-intensive technology that slows down the performance when the system is overloaded. Issues might occur with a user action or with the quality of the analytics. Despite the general tendency for low power consumption, advanced analytics requires a powerful CPU or implementation of an ASIC or SoC specialized for AI computing with low energy consumption.
Power consumption optimization
When it comes to enterprise IoT applications, all methods for decreasing the power consumption within PCB work here, as IoT processing and analytical tools are incredibly energy-consuming. For instance, sleep mode for the unused module decreases the power consumption within a particular application, while the utilization of switching regulators helps prolong the battery life. When choosing between standard and low-energy interfaces, without a doubt, choose the latter. Bluetooth Low Energy (BLE) and Zigbee might help here.
To balance the power consumption, it would be effective to provide a permanent alternative energy source. The energy can be additionally harvested through thermoelectric, electromagnetic, radio frequency, piezoelectric, photovoltaic and other methods.
Accessibility and durability
The need to serve difficult-to-reach areas affirms the benefit of multiple wireless protocols within one board. For this purpose, companies might go with BLE mesh networks that also overcome obstacles like walls and tunnels. It can connect more devices than other mesh nets, but increasing the volume of data transmitted can cause delays. Thus, every protocol has its advantages and restrictions, but by providing more options, you provide more potential solutions for the enterprise.
The other point associated with the environment of use is that IoT devices often require data collection from moving or vibrating objects, which induces advanced protection for such systems. This is especially true for logistics and automotive industries. In addition to the standard recommendation of how to protect the board, it is worthy to utilize anti-vibrational frames. Even though it increases the cost of the final solutions, it increases the chances for the data to be completely collected even after an accident.
Where can EIoT PCB design harness the experience of consumer IoT?
Enterprise IoT is not only analytical systems or advanced manufacturing equipment but also wearables and HMI-enabled devices when it comes to interaction with people. The rich experience of the consumer electronics industry can and should be applied to the development of IoT ecosystems at enterprises that interact with people, such as healthcare or worker safety applications.
The main factor for consumer IoT devices is the ease of their use, which affects the PCB design first. It means the following requirements must be met for PCB design:
- Compactness. For now, to provide the small size of the device, developers can resort to flex and high-density–interconnect PCBs that also allow them to adapt to the form and shape of the future device.
- Noiseless. It is required for seamless communication between devices first. Any electrical or reflection noise needs to be eliminated within the PCB, which may require the use of noise filters and damping resistors.
- Durability. To ensure the device is durable, there is a need to simulate conditions of use and implement suitable compensation schemes.
In contrast to the consumer, which involves connecting the wearable mainly to the phone, there might be a need for EIoT wearables to transmit information to other devices, to the control panel and to the cloud for further analysis. Also, more complex equipment is designed to collect heterogeneous data about human health. Thus, this reinforces the issue of multi-functionality for the small device.
- When designing PCB for EIoT, consider high-level benefits and industry trends, like enhanced requirements for the quality of analytics, production flexibility, and multi-purpose IoT.
- The main task for the PCB designer of enterprise IoT applications is to strike a balance between multi-functionality and low energy consumption.
- Consider technological innovations for more computing capabilities of AI applications.
- Utilize the experience from consumer IoT to increase usability for human-oriented applications, like healthcare or worker safety IoT solutions.