SAFETY CRITICAL EMBEDDED SYSTEMS & AI

Ferguson Control Systems (FCS), LLC is a Female Minority Owned Business based on the experience of Roscoe C. Ferguson Jr., a former IBMer and renown NASA contractor.

An embedded system is a computer system—a combination of a computer processor, computer memory, and input/output peripheral devices—that has a dedicated function within a larger mechanical or electronic system. It is embedded as part of a complete device often including electrical or electronic hardware and mechanical parts. Because an embedded system typically controls physical operations of the machine that it is embedded within, it often has real-time computing constraints. Embedded systems control many devices in common use.

Founded in 2012, Ferguson Control Systems is an innovative and proven company that specializes in the architecture, design, and implementation of embedded systems software. Located in Houston, Texas, the company is composed of experienced developers of on-board systems from Human Space Flight and Military Programs. FCS has embraced the knowledge learned from the development of human-rated and safety critical systems and utilizes these techniques to provide quality and reliability in the development of control systems software.

The Apollo Guidance Computer was one of the first modern embedded systems. It was developed around 1965 by Charles Stark Draper at the MIT Instrumentation Laboratory. The computer was considered risky for the Apollo project because it used newly developed monolithic integrated circuits to make it smaller and lighter.

The Autonetics D-17 guidance computer for the Minuteman missile was one of the first embedded systems to be mass-produced. It was released in 1961. In 1966, when the Minuteman II began production, a new computer was introduced to replace the D-17. This new computer was significant because it marked the first time integrated circuits were used in large quantities.

Embedded systems have experienced a significant decrease in cost since their initial applications in the 1960s. Additionally, there has been a noteworthy advancement in processing power and functionality. The Intel 4004, an early microprocessor introduced in 1971, was primarily designed for calculators and other small systems. However, it still required external memory and support chips. In the early 1980s, the integration of memory, input, and output system components into the same chip as the processor led to the development of microcontrollers. Microcontrollers are utilized in scenarios where utilizing a general-purpose computer would be excessively expensive. As the cost of microprocessors and microcontrollers decreased, the utilization of embedded systems significantly increased.

A microcontroller with a relatively low cost can be programmed to perform the same functions as multiple separate components. The use of microcontrollers has made it possible to replace expensive analog components like potentiometers and variable capacitors with up/down buttons or knobs that are read by a microprocessor, even in consumer products. While an embedded system is typically more complex than a traditional solution in this context, the majority of the complexity resides within the microcontroller itself. Only a few additional components may be required, and the main focus of the design process is on the software. In comparison to designing and building a new circuit without an embedded processor, prototyping and testing the software can be done more quickly.

Ferguson Control Systems specializes in the architecture, design, and implementation of embedded systems software.

Network Logic Programming Theory

Logic abstracted as relationships and their derivates is the next evolution of programming. There is an assumption that, with the rise of artificial intelligence, the art of programming is dead, in part because it has the reputation of being expensive, with costs rising and production timelines expanding. But can anything be done? The answer is yes. However, complexity must be better understood.

This book introduces Network Logic Programming Theory, which tackles programming’s challenges with a technique that separates complex programming algorithms into networks and computations (U.S. Patent 12,131,160 and Patent Pending). Networks store logic in the form of relationships to make decisions, while computations are reduced to simple algorithms. This network-based system utilizes the principles of network science to analyze complexity within software systems. AI and Network Logic Programming Theory are based on the concept of relationships and their derivates. The difference is AI extracts relationships from data using networks, while Network Logic Programming Theory programs networks using relationship-based abstraction. Such an approach can serve as the next evolution of programming, and be integrated with AI to create deterministic systems. Networks become a new version of assembly language and can store both static and dynamic information for data processing algorithms.

The use of network logic makes it more natural to open the door for concepts such as runtime verification, true distributed logic processing systems (including network logic distributed across geographically dispersed locations), or more deterministic adaptive systems by providing generative logic paths. Generative logic paths allow for the runtime generation of logic using predefined relationships such as the capability of a GPS Application that calculates routes based on a predetermined set of streets and roads (relationships).

When coupled with AI, the opportunities are to use Network Logic Programming Theory to support deterministic decision making, use AI models that learn relationships from data to produce static graphs that define these relationships as input to Network Logic Programming Theory network programs, and have AI generate systems using Network Logic Programming Theory so the generated complexity can be understood using Network Science technology. All in the name of supporting deterministic systems.

Network Logic Programming Theory introduces the concept of the Network Processor. The Network Processor is designed to process network flow logic for static logic paths defined in network definitions. It can also generate new paths dynamically during run time based on stored relationships. It can respond to queries during runtime about how items are related similar in concept to the query capability of a graph database such as Neo4j using cypher. The Network Processor and CPU work together to support the data processing model.