Mitch Fletcher

Next generation space avionics: a highly reliable layered system implementation

Advances in electronics over the past decade have produced major improvements in the power and flexibility of personal computer systems. Unfortunately current avionics systems for space applications typically have not leveraged these COTS advantages. Recently, there has been a trend toward utilization of commercial bus interconnects, primarily VME and PCI. These parallel interconnects have the disadvantage that a single failure can disable an entire string of the redundancy. Honeywell has developed a patent pending architecture for an avionics system that combines the high reliability of previous serial systems with the flexibility and openness of direct COTS bus interface. A decade ago, the state of the art for avionics systems made a step change from the PAVE PILLAR systems of the 1980s to the integrated modular avionics (IMA) used in the Boeing 777. This next generation avionics architecture is not based upon traditional Byzantine redundancy structures, but on a truth based scheme where each element knows when an internal failure occurs and removes itself from the system. IMA utilizes a lock step microprocessor design that communicates to a COTS backplane for input/output, and to a virtual backplane/sup /spl trade// (a reliable, high-speed serial bus such as 1394 or AFDX) for intra-system communication. The system functions are implemented using an ARINC-653 time and space partitioned operating system. The entire system provides the simplicity of a simplex system, implements the highest level of reliability provides complete flexibility to reconfigure both software applications and hardware interfaces, allows for rapid prototyping using low-cost COTS hardware, and is easily expandable beyond the initial point implementation. As the only 5th generation avionics architecture, the concepts incorporated into Honeywells IMA are ideally suited to be the backbone of the next generation Crew Exploration Vehicle for Project Constellation.

Simplified Robotics Avionics System: A Integrated Modular Architecture Applied Across a Group of Robotic Elements

The latest NASA initiative for Human Space, namely the Space Exploration Vision, which encompasses Project Constellation, provides new opportunities for system implementation. The second wave of development after Crew Exploration Vehicle and Crew Launch Vehicle development, and following Shuttle retirement, will be development of lunar base concepts and operations leading to early robotic missions. The current vision for lunar base implementation anticipates that there will be highly integrated robotic pre-construction operations and robotic assistants for the astronauts. In preparation for this robotics involvement, there is a series of robotic precursor missions to the Moon and Mars. Historically, many humans are required to control a single robot; in practice the Mars Exploration Rovers require a staff of approximately 70 to support continual operation of a single robotic rover. In addition robotic avionics has typically been customized for each robot. While this has been effective for prior robotic missions, the habitation and exploration of the Moon and Mars requires many robots working in tandem with humans. The limited NASA budget to implement the Space Exploration Vision requires that multiple robots be commanded by a minimal operations staff and that a common set of avionics electronics be used across the multitude of robots needed. Traditional robotic avionics do not address either the additional autonomy or commonality required by this new set of robotic missions. One solution to address these concepts is to apply a Honeywell patent pending architecture that uses an integrated modular avionics (IMA) approach across a multiplicity of robots. This concept treats a group of robotic elements as a single system. Instead of each robot having a separate avionics system, a single shared avionics system will be deployed across the robots. This sharing would be implemented using an IMA system approach with each element of the robotic system being connected using a Virtual Backplanetrade. The IMA approach is a next generation avionics architecture where each element knows when an internal failure occurs and removes itself from the system. IMA utilizes a fail passive design that communicates to a COTS backplane for input/output and to the aforementioned Virtual Backplanetrade for intra-system communication. Each robot implements either single or multiple hardware-enhanced ARINC-653 software partitions. Together these partitions form a single system that provides the simplicity of a simplex system; implements the highest levels of reliability; provides the flexibility to easily reconfigure both software applications and hardware interfaces; allows for rapid prototyping using low-cost COTS hardware; and is easily expandable beyond the initial point implementation. The avionics for each robot interfaces to the local sensors and effectors. The high-level control of the robot may be local or may reside on another robot, a group of robots, or a remote base station. From a system standpoint, control of multiple robots is viewed as a single system with multiple components as opposed to multiple individual systems interacting together. The system level control could include redundant elements spread across multiple robots depending on the level of fault tolerance and reliability that is required. The robotic system could also be dynamically reconfigured when multiple elements (robot assistants, robotic vehicles) join or leave the system, adjusting to changing mission needs. The application of IMA principles to robotics applications provides an infrastructure that has been demonstrated to reduce cost, schedule, and risk throughout the life of the program. In addition, this infrastructure provides the means for applying new approaches to solving problems such as multi-robot collaboration

IMA Readiness for Autonomous Robotic Systems in Extraterrestrial Surface Exploration Missions

Recent developments in hardware readiness are closing the technological gaps which limited the implementation of robotic autonomy in extraterrestrial surface exploration missions. Innovative applications of wireless technology and avionics architectural principles drawn from the Orion crew module, to name one example, provide solutions for several of these gaps. If future space exploration missions are to grow significantly more complex, greater levels of autonomy must be afforded to robotic systems. This paper describes how Orion’s avionics architecture attributes can be leveraged to implement an independent, deterministic “Safety” partition” that prevents non-deterministic autonomous applications from issuing unsafe commands. The certification issues endemic to having autonomous applications operating alongside humans is also addressed. Robust avionic architectures by themselves are not sufficient, requiring aggressive innovations in size, weight and power, to allow the avionics hardware infrastructure to meet stringent robotic mission requirements. The emerging next generation of integrated modular avionics addresses this challenge with smaller, but very capable platforms. Another technology gap being addressed through the use of avionics architectural principles and deterministic wireless technology is the coordination of multiple autonomous systems. As a proof-ofconcept, Honeywell has developed various algorithms and wireless hardware that lead to a deterministic, fault tolerant, reliable wireless backplane. Honeywell has developed a laboratory facility based upon the Simulation & Modeling for Acquisition, Requirements and Training (SMART) concepts. Through this SMARTlabTM facility multi-robot collaboration can be achieved via interaction of real and simulated robots, rather than requiring the presence of several costly, physical robots. By filling technology gaps associated with space based autonomous systems, recent advances in wireless and simulation technology, along with Orion architectural principles, provide the means for decreasing operational costs and simplifying problems associated with autonomous systems, including those requiring the collaborative work of multiple robotic assets.