Access Official Technical PDF

Share Summary:LinkedInEmail

Semiconductor startups are working on technologies that could address several persistent industrial automation challenges: moving data reliably from connected devices, processing information close to where it is generated, protecting embedded systems and managing power in battery-operated equipment. A Fall 2019 update from Silicon Catalyst, an incubator focused on semiconductor solution startups, offered a snapshot of companies pursuing these areas. Their work illustrates how developments in silicon, sensors and communications can intersect with industrial IoT (IIoT) system design.

The portfolio included companies working on wireless infrastructure, edge computing, embedded security, energy efficiency and sensing. These were distinct approaches rather than a single integrated platform, but together they point to design considerations for engineers, system integrators and operations leaders evaluating connected equipment.

Semiconductor startups address IIoT connectivity and coverage

Industrial IoT deployments depend on communications links suited to the application, site and device. The portfolio featured approaches ranging from long-distance wireless transport to short-range, low-power connections.

Beam Semiconductor described a 28 GHz and 60 GHz wireless transport solution aimed at applications including cellular backhaul, fiber extension and fixed wireless access. Its proposed system combined RF silicon with a phased-array MicroHorn antenna. The company said analog beam steering could automatically shape and direct the signal, reducing the need for manual antenna positioning. For industrial and infrastructure projects, the underlying design challenge is familiar: extending high-capacity connectivity to locations where installing fiber is costly, slow or impractical.

SPARK Microsystems focused on short-range ultra-wideband wireless for connected devices. Its stated applications included industrial automation and machine-to-machine communications. The company positioned its radio around low power use, reduced latency and scalable data rates, with the potential to support battery-operated or energy-harvesting sensor nodes. These characteristics matter when evaluating wireless links for distributed sensors, although deployment choices still need to reflect each system’s range, data and reliability requirements.

Espre Technologies presented another approach, combining communications, sensor and FPGA-based products for military and commercial IIoT applications. Its stated aims included improving connectivity and managing interference through signal processing and AI-based features. The portfolio’s range of approaches underscores that industrial wireless is not one problem: a plant may need local sensor connectivity, site-wide coverage or a backhaul link, each with different constraints.

Edge computing brings processing closer to equipment

Connected machines can generate data that must be interpreted quickly or handled without relying on a distant computing resource. Edge computing addresses this by placing processing nearer to sensors and devices. The Silicon Catalyst update included companies developing processors and architectures for this purpose.

Quadric described a unified processor architecture intended to run parallelizable workloads that might otherwise be spread across separate components such as CPUs, GPUs, FPGAs and AI accelerators. Its stated objective was to simplify the hardware and software approach for edge applications. For integrators, the broader issue is how to balance compute performance, latency, power consumption and development complexity when selecting edge hardware.

Mentium Technologies presented Analog Computing Memory, which uses non-volatile memory devices to store neural-network parameters and perform computation. The company’s concept aimed to reduce the need to move data between memory and processing resources. This approach reflects a key challenge in edge AI design: the movement of data can be an important part of a system’s computing and energy demands.

Embedded security and power management shape device design

Cybersecurity for industrial equipment is not limited to network defenses. Dover Microsystems described CoreGuard, a silicon-level approach designed to monitor processor instructions against defined security, safety and privacy rules, called micropolicies. According to the company, a violating instruction would be stopped before execution. The proposal highlights the value of considering security at the processor level, alongside software controls and communications protections, particularly for embedded devices that may operate for long periods.

Power is another foundational design constraint, especially for wireless sensors and portable equipment. EcoCircuits described a Battery Booster chip intended to manage the trade-off between battery life and the peak power needed for wireless transmission. Power Down Semiconductor presented Low Power Processing technology designed to recycle charge in circuit elements including memory blocks, clock buffers, GPIO and displays. Both examples target energy use at the silicon or circuit level—an important consideration where battery replacement, device size or energy harvesting affects deployment and maintenance planning.

Sensing advances can support richer machine data

Industrial systems rely on sensors to turn physical conditions into usable information. SigmaSense focused on digital approaches to analog sensing, including touch sensing for displays, while Probius Dx described an electronic platform for analyzing biological samples. These applications differ from conventional factory instrumentation, but they demonstrate the breadth of sensor-related semiconductor innovation across industrial, medical and other markets.

For automation teams, the practical lesson is to assess sensing technologies against the required signal quality, environmental conditions, integration effort and data-processing needs. A sensor’s value depends not only on the measurement itself, but also on how reliably its output can be interpreted and used by the wider control or information system.

What industrial teams should take from the portfolio

The Fall 2019 update is a historical snapshot, not a current product availability guide. Its value for industrial decision-makers is the set of engineering themes it brings together:

  • Match connectivity to the use case: distinguish between local sensor links, wide-area coverage and high-capacity wireless transport.
  • Evaluate edge processing as a system choice: consider compute architecture, software workflow, power and latency together.
  • Include security in hardware discussions: assess embedded protections as part of a layered cybersecurity strategy.
  • Account for energy constraints early: battery life and peak transmission power can influence device size and maintenance requirements.
  • Plan for integration: new silicon, sensing methods and communications technologies must fit the application’s existing equipment and operating environment.

For plant managers, system integrators and technical leaders, the broader implication is that semiconductor innovation can affect industrial automation well beyond the controller or PLC. Progress in connectivity, edge computing, embedded security, sensing and power management can shape how IIoT devices are designed and deployed. Evaluating these technologies against the needs of a specific operation—rather than relying on broad performance claims—is essential to turning promising chip-level ideas into practical automation systems.

Access Official Technical PDF

About The Author