The Internet of Things (IoT) has transformed the way products collect data, communicate, and interact with the world around them. From industrial equipment and aerospace systems to smart devices and connected vehicles, IoT products increasingly depend on continuous streams of data to operate efficiently and intelligently.
However, as connected products generate more data, relying entirely on cloud computing can create challenges related to latency, connectivity, bandwidth, security, and scalability. This is where edge computing becomes an important part of modern IoT product design.
By processing data closer to where it is generated, edge computing can help IoT products respond faster, reduce dependence on remote infrastructure, and operate more effectively in environments where connectivity may be limited. For engineering organizations such as SunMan Engineering, incorporating edge computing considerations early in the product development process can create opportunities to develop more responsive, reliable, and scalable connected products.
Edge computing is a distributed computing approach in which data is processed closer to the physical location where it is generated rather than sending every piece of data to a centralized cloud or data center.
In a traditional IoT architecture, sensors and devices collect information and transmit it to a cloud platform for processing and analysis. The cloud then sends instructions or results back to the device.
With edge computing, some of that processing occurs directly on the IoT device or on a nearby gateway, industrial computer, embedded processor, or other edge device.
A simplified architecture looks like this:
Instead of sending all raw data to the cloud, the edge device can filter, analyze, aggregate, or act on the data locally. Only the information that needs centralized storage, advanced analytics, or broader system coordination needs to be transmitted to the cloud.
Edge computing is not simply a software decision. It can influence the architecture of an IoT product from the earliest stages of development.
Product designers and engineers need to consider where data should be collected, where it should be processed, how quickly the system needs to respond, and what happens when network connectivity is unavailable.
Several factors make edge computing particularly valuable.
One of the biggest advantages of edge computing is reduced latency.
If an IoT device must send sensor data to a remote cloud server, wait for processing, and then receive a response, even a small network delay can affect system performance.
For applications requiring near-real-time decisions, local processing can be significantly more practical.
For example, an industrial monitoring system could analyze vibration, temperature, or pressure data locally and immediately identify an abnormal operating condition. The system does not necessarily need to wait for a cloud application to process every individual sensor reading.
This can be especially important for applications involving equipment monitoring, automation, robotics, transportation, and other systems where timely responses matter.
IoT products may generate enormous quantities of data.
Sending every sensor measurement continuously to the cloud can increase bandwidth consumption and infrastructure costs. Edge computing allows the device to process information locally and transmit only meaningful results.
For example, instead of transmitting thousands of raw sensor measurements, an edge device could send:
This approach can make the overall IoT architecture more efficient.
Many IoT products operate in locations where reliable network connectivity cannot be guaranteed.
Industrial facilities, remote monitoring systems, transportation applications, aerospace environments, and field-deployed equipment may experience intermittent or restricted connectivity.
An edge-enabled product can continue performing important functions locally even when its connection to the cloud is interrupted.
Once connectivity is restored, the system can synchronize relevant information with the cloud.
This capability can be an important consideration when designing products that must maintain operational continuity.
Not all IoT data has the same value.
A sensor might generate thousands of readings, but only a small percentage may be relevant for long-term storage or further analysis. Edge processing provides an opportunity to determine which information matters before transmitting it.
Engineers can design systems that distinguish between:
This can reduce unnecessary data transmission while allowing critical information to reach cloud-based systems and enterprise applications.
Processing information locally can also provide additional opportunities for controlling sensitive data.
Depending on the application, certain information may not need to leave the device or local network. Edge computing can allow designers to perform selected analysis locally and transmit only the necessary results.
Security, however, should not be viewed as an automatic benefit of edge computing. Distributed computing also creates additional endpoints that must be protected.
IoT product designers should therefore consider:
Security needs to be incorporated into the architecture rather than added after the product has already been designed.
The relationship between edge computing and embedded systems is particularly important in IoT product development.
Many IoT products already contain microcontrollers, microprocessors, sensors, communication modules, and embedded software. Adding edge intelligence can allow these systems to perform increasingly sophisticated processing locally.
Depending on the application, an edge-enabled embedded system may perform:
The appropriate hardware depends on the computational requirements of the product.
Engineers may need to evaluate processor performance, memory requirements, power consumption, thermal constraints, communication interfaces, operating systems, and software architecture when determining how much processing should occur at the edge.
One of the most significant developments associated with edge computing is the integration of artificial intelligence and machine learning into IoT products.
Traditional AI architectures often rely heavily on cloud computing. Edge AI moves some machine-learning inference closer to the source of the data.
For example, an intelligent industrial device could use a trained model to recognize unusual vibration patterns directly on the device. Rather than transmitting every waveform to the cloud for analysis, the device can identify potentially important events locally.
This creates opportunities for:
For product developers, this means AI considerations increasingly need to become part of the hardware and software architecture rather than being treated as a separate cloud-based feature.
There is no single architecture that works for every IoT product.
Some products may perform almost all processing locally. Others may use a hybrid architecture in which edge devices handle time-sensitive processing while the cloud handles large-scale analytics, data storage, visualization, and machine-learning model management.
A practical design process should consider several questions:
Time-sensitive data should generally be evaluated for local processing requirements.
Historical information may be better suited for centralized cloud or enterprise systems.
Processing requirements need to be balanced against cost, power consumption, thermal performance, and physical constraints.
The product should have a clearly defined operating strategy for offline or degraded network conditions.
Edge devices require mechanisms for secure firmware, software, and model updates throughout their operational lifecycle.
Security must be considered across hardware, firmware, software, communications, data, and cloud interfaces.
Edge computing affects more than the final electronics design. It can influence the entire product realization process.
A successful IoT product requires coordination between multiple engineering disciplines, including:
The decisions made during product architecture can determine how effectively these disciplines work together.
This is where an integrated engineering approach becomes valuable.
SunMan Engineering brings together electronic and mechanical engineering, embedded software and firmware development, product architecture, prototyping, and product realization capabilities.
For IoT products incorporating edge computing, these capabilities can help organizations address the relationship between hardware, embedded intelligence, communications, and the broader product architecture.
Rather than treating the IoT device, firmware, and mechanical product as independent components, an integrated development approach considers how they work together as a complete system.
For example, an edge-enabled product may require decisions involving:
Hardware:
Selecting processors, memory, sensors, communication interfaces, and power architecture.
Firmware:
Developing the embedded software responsible for sensor acquisition, local processing, device control, and communications.
Mechanical Design:
Managing enclosure requirements, thermal considerations, environmental protection, and physical integration.
Connectivity:
Selecting and integrating appropriate communication technologies for the intended operating environment.
Data Architecture:
Determining what information is processed locally, what is transmitted, and how edge and cloud systems interact.
Prototyping and Validation:
Testing the complete system under realistic operating conditions before moving toward production.
This integrated perspective can help reduce architectural disconnects and identify engineering challenges earlier in the development cycle.
Allen Nejah and a System-Level Approach to IoT Product Development
At the intersection of product strategy, engineering, and technology development, Allen Nejah brings a system-level perspective to complex product development initiatives.
For IoT products, this perspective is particularly relevant because successful products require more than connecting sensors to a network. Teams need to understand how technology decisions support product requirements, business objectives, manufacturing realities, scalability, and the end-user experience.
Edge computing provides an excellent example of why these considerations need to be addressed early.
The question is not simply:
The more important questions are:
A system-level product development approach helps ensure that edge computing is implemented because it solves a meaningful product or operational requirement—not simply because it is a current technology trend.
As IoT products become more intelligent, the distinction between connected devices and computing platforms continues to disappear.
Modern connected products are increasingly expected to sense their environment, interpret information, make decisions, communicate with other systems, and respond with minimal delay.
Edge computing provides an architectural foundation for this evolution.
Future IoT products will likely combine:
Sensors + Embedded Processing + Edge AI + Connectivity + Cloud Computing + Data Analytics
The cloud will continue to play an important role in centralized analytics, fleet management, storage, model development, and enterprise integration. At the same time, edge devices will increasingly handle time-sensitive processing and decision-making closer to the source.
Edge computing is becoming an important consideration in the design of modern IoT products. By moving selected processing closer to where data is generated, organizations can potentially improve responsiveness, reduce network requirements, support operation in disconnected environments, and build more intelligent connected products.
However, implementing edge computing effectively requires careful consideration of the entire product architecture.
From hardware and embedded firmware to mechanical design, connectivity, security, cloud integration, and lifecycle management, every element needs to work together.
For companies developing sophisticated IoT products, SunMan Engineering can provide an integrated engineering and product development perspective that connects these disciplines from concept through prototyping and product realization.
With the right architecture and engineering strategy, edge computing can move IoT beyond simple connectivity—helping create products that can sense, analyze, decide, and act closer to where the real-world data is generated.
What our clients say
Established in 1990, SunMan Engineering has engaged and assisted over 1550 leading technology companies in successfully completing over 1664 product development projects to date.