Failure determination system

Abstract

A system-failure determination system fo a vehicle, made up of processors linked together that control or monitor the systems of a vehicle, that includes critical and non-critical applications that generate data. Certain data from the critical applications would be useful for the reconstruction of a system failure or for the reconstruction of an accident; and configuring an executive to monitor and store data for later recovery to determine cause.

Description

BACKGROUND OF THE INVENTION

Java is a robust, object-oriented programming language expressly designed for use in the distributed environment of the Internet. Java can be used to create complete applications that may run on a single computer or be distributed among servers and clients in a network. A source program in Java is compiled into byte code, which can be run anywhere in a network on a server or a client that has a Java virtual machine (JVM).

A JVM describes software that is nothing more than an interface between the compiled byte code and the microprocessor or hardware platform that actually performs the program's instructions. Thus, the JVM makes it possible for Java application programs to be built that can run on any platform without having to be rewritten or recompiled by the programmer for each separate platform.

Jini is a distributed system based on the idea of federating groups of users and the resources required by those users. Resources can be implemented either as hardware devices, software programs, or a combination of the two. The Jini system extends the Java application environment from a single virtual machine to a network of machines. The Java application environment provides a good computing platform for distributed computing because both code and data can move from machine to machine. The Jini infrastructure provides mechanisms for devices, services, and users to join and detach from a network. Jini systems are more dynamic than is currently possible in networked groups where configuring a network is a centralized function done by hand.

However, the Java/Jini approach is not without its disadvantages. Both Java and Jini are free, open source applications. The Java application environment is not designed for controlling messaging between different machines. For example, the Java application is not concerned about the protocols between different hardware platforms. Jini has some built-in security that allows code to be downloaded and run from different machines in confidence. However, this limited security is insufficient for environments where it is necessary to further restrict code sharing or operation sharing among selected devices in a secure embedded system.

SUMMARY OF THE INVENTION

The present invention allows construction of a secure, real-time operating system from a portable language such as Java that appears to be a Java virtual machine from a top perspective but provides a secure operating system from a bottom perspective. This allows portable languages, such as Java, to be used for secure embedded multiprocessor environments.

The foregoing and other objects, features and advantages of the invention will become more readily apparent from the following detailed description of a preferred embodiment of the invention which proceeds with reference to the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a diagram showing a java stack with an additional Secure Real-time Executive (SRE) layer.

FIG. 2 is a diagram of a multiprocessor system that runs multiple Java Virtual Machines that each include a SRE.

FIG. 3 is a detailed diagram of the managers in the SRE.

FIG. 4 is a block diagram of how the SRE manages a multiprocessor system.

FIG. 5 is a block diagram showing how a task manager in the SRE operates the multiprocessor system in a lock-step mode.

DETAILED DESCRIPTION

A java application stack includes a Java layer 5 for running any one of multiple different applications. In one example, the applications are related to different vehicle operations such as Infrared (IR) and radar sensor control and monitoring, vehicle brake control, vehicle audio and video control, environmental control, driver assistance control, etc. A Java Virtual Machine (JVM) layer 16 provides the hardware independent platform for running the Java applications 5. A Jini layer 12 provides some limited security for the Java applications that run on different machines. However, the Jini layer 12 does not provide the necessary reconfiguration and security management necessary for a distributed real-time multiprocessor system.

A Secure Real-time Executive (SRE) 14 provides an extension to the JVM 16 and allows Java to run on different processors for real-time applications. The SRE 20 manages messaging, security, critical data, file I/O multiprocessor task control and watchdog tasks in the Java environment as described below. The JVM 16, Jini 12 and SRE 14 can all be implemented in the same JVM 10, However, for explanation purposes, the JVM 10 and the SRE 14 will be shown as separate elements.

FIG. 2 shows a system 15 that includes multiple processors 16, 18, 20, 22 and 24. Each processor includes one or more JVMs 10 that run different Java applications. For example, processor 16 includes one Java application 28 that controls a vehicle security system and another Java application 26 that controls the vehicles antilock brakes. A processor 18 includes a Sava application 30 that controls audio sources in the vehicle, Other processors 20 and 22 may run different threads 32A and 32B for the same sensor fusion Java application 32 that monitors different IR sensors. Another thread 32C on processor 24 monitors a radar sensor for the sensor fusion Java application 32.

The SRE 14 runs below the JVMs 10 in each processor and control tasks, messaging, security, etc. For example, the Java application 26 controls vehicle braking according to the sensor data collected by the sensor fusion Java application 32. The SRE 14 in one example prevents unauthorized data from being loaded into the processor 16 that runs brake control application 26. The SRE 14 also prevents other Java applications that are allowed to be loaded into processor 16 from disrupting critical braking operations, or taking priority over the braking operations, performed by Java application 26,

For example, the SRE 14 may prevent noncritical vehicle applications, such as audio control, from being loaded onto processor 16. In another example, noncritical operations, such as security control application 28, are allowed to be loaded onto processor 16. However, the SRE 14 assigns the security messages low priority values that will only be processed when there are no braking tasks in application 26 that require processing by processor 16.

The SRE 14 allows any variety of real-time, mission critical, nonreal-time and nonmission critical Java applications to be loaded onto the multiprocessor system 15, The SRE 14 then automatically manages the different types of applications and messages to ensure that the critical vehicle applications are not corrupted and processed with the necessary priority. The SRE 14 is secure software that cannot be manipulated by other Java applications.

The SRE 14 provides priority preemption on a message scale across the entire system 15 and priority preemption on a task scale across the entire system 15, So the SRE 14 controls how the JVMs 10 talk to each other and controls how the JVMs 10 are started or initiated to perform tasks. The SRE 14 allows programmers to write applications using Java in a safe and secure real time environment, Thus, viruses can be prevented by SRE 14 from infiltrating the system 15.

While the explanation uses Java as one example of a programming environment where SRE 14 can be implemented, it should be understood that the SRE 14 can be integrated into any variety of different programming environments that may run in the same or different systems 15. For example, SRE 14 can be integrated into an Application Programmers Interface (API) for use with any programming language such as C++.

FIG. 3 shows the different functions that are performed by the SRE 20. Any combination of the functions described below can be provided in the SRE 20. A message manager 50 controls the order messages are received and transmitted by the different Java applications. A security manager 52 controls what data and messages are allowed to be received or transmitted by different Java applications. A critical data manager 54 controls what data is archived by the different Java applications.

A data manager 56 controls what data is allowed to be transferred between different processors. A task manager 58 controls the order tasks are performed by the different JVMs. A reconfiguration manager 60 monitors the operation of the different processors in the system and reassigns or reconfigures Java applications and Java threads to different processors according to what processors have failed or what new processors and applications have been configured into system 15.

The message manager 50 partially corresponds to the priority manager 44 shown in FIG. 2 of U.S. Pat. No. 6,629,033, issued Sep. 30, 2003, the critical data manager 52 partially corresponds with the logging manager 44 shown in FIG. 2 of the '033 patent, and the security manager 54 at least partially corresponds with the security manager 40 shown in the '033 patent. The data manager 56 at least partially corresponds with the date manager 42 shown in FIG. 2 of U.S. Pat. No. 7,146,260 issued Dec. 5, 2006, the task manager 58 partially corresponds to the device manager 46 shown in FIG. 2 of the '260 patent, and the configuration manager 60 at least partially corresponds to the configuration manager 44 shown in FIG. 2 of the '260 patent. The descriptions of how the different manager 50-60 operate similarly to the corresponding manager in the '033 and '260 patents are herein incorporated by reference and are therefore not described in further detail.

However, some specific tasks performed by the managers 50-60 are described below in further detail.

FIG. 4 shows in more detail how the SRE 14 operates, One of the operations performed by the task manager 58 is to control when different tasks are initiated on different processors. For example, a first Global Positioning System (GPS) thread 62 is running on a JVM in a processor 80. Another sensor fusion thread 64 is running on a different processor 82. Block 74 represents the Java Virtual Machine operating in each of processors 80 and 82. A master JVM 74 may run on either processor 80, processor 82 or on some other processor.

The task manager 58 sends an initiation command 66 to the GPS thread 62 to obtain location data. The task manager 58 then directs the obtained GPS data 68 through a link to the sensor fusion thread 64 for subsequent processing of GPS data 68. The link may be any bus, such as a PCI bus, serial link such as a Universal Serial Bus, a wireless link such as blue tooth or IEEE 802.11, or a network link such as Ethernet, etc.

The configuration manager 60 acts as a watchdog to make sure that the GPS thread 62 and the sensor fusion thread 64 are each running correctly. In one example, separate configuration managers 60 in each processor 80 and 82 sends out periodic signals to the other configuration managers 60 in the other processors, Any one of the configuration managers 60 can detect a processor or application failure by not receiving the periodic “ok” signals from any one of the other processors for some period of time, If a failure is detected, then a particular master configuration manager 60 in one of the processors determines where the task in the failed processor is going to be reloaded. If the master configuration manager 60 dies, then some conventional priority scheme, such as round robin, is used to select another configuration master.

If a failure is detected, say in the processor 82 that is currently performing the sensor fusion thread 64, a message is sent from the configuration manager 60 notifying the task manager 58 which processor is reassigned the sensor fusion thread. In this example, another sensor fusion thread 76 in processor 84 is configured by the configuration manager 60.

The critical data manager 52 manages the retention of any critical data 72 that was previously generated by the sensor fusion thread 64. For example, the critical data manager 54 automatically stores certain data and state information that was currently being used in the sensor fusion thread 64. The critical data may include GPS readings for the last 10 minutes, sensor data obtained from sensors in other processors in the vehicle over the last 10 minutes. The critical data may also include any processed data generated by the sensor fusion thread 64 that identifies any critical vehicle conditions.

The critical data manager 52 also determines which data to archive generally for vehicle maintenance and accident reconstruction purposes. The configuration manager 60 directs the critical data 72 to the new sensor fusion thread 76. The task manager 74 then redirects any new GPS data obtained by the GPS thread 78 to the new sensor fusion thread 76 and controls sensor fusion tasks from application 76. Thus, the configuration manager 60 and the task manager 58 dynamically control how different Java threads are initialized, distributed and activated on different processors.

The message manager 50 determines the priority of sent and received messages. If the data transmitted and received by the sensor fusion thread 76 is higher priority than other data transmitted and received on the processor 84, then the sensor fusion data will be given priority over the other data. The task manager 58 controls the priority that the sensor fusion thread 76 is giving by processor 84. If the sensor fusion thread 76 has higher priority than, for example, an audio application that is also being run by processor 84, then the sensor fusion thread 76 will be performed before the audio application,

The SRE 14 can be implemented in any system that needs to be operated in a secure environment. For example, network servers or multiprocessors operating in a home environment. The multiprocessors in home appliances, such as washer and dryers, home computers, home security systems, home heating systems, can be networked together and operate Java applications. The SRE 14 prevents these multiple processors and the software that controls these processors from being corrupted by unauthorized software and also allows the applications on these different processors to operate as one integrated system,

The SRE 14 is a controlled trusted computing based that is not accessible by non-authorized application programmers and anyone in the general public. Therefore, the SRE 14 prevents hacking or unauthorized control and access to the processors in the vehicle.

Task Controlled Applications

Debugging is a problem with multiprocessor systems. The task manager 58 allows the Java applications to be run in a lock-step mode to more effectively identify problems in the multiprocessor system 15.

FIG. 5 shows a path 90 taken by a vehicle 92. In one application, the position of the vehicle 92 is sampled every second t₁, t₂, t₃, t₄, etc. The position of the vehicle 92 is sampled by a GPS receiver in vehicle 92 that reads a longitudinal and latitudinal position from a GPS satellite. The UPS receiver is controlled by the GPS thread 62 that receives the GPS data and then sends the GPS data to a sensor fusion thread 64 that may run on the same or a different processor in the vehicle 92. The sensor fusion thread 64 can perform any one of many different tasks based on the GPS data. For example, the sensor fusion thread 64 may update a map that is currently being displayed to the driver of vehicle 92 or generate a warning signal to the vehicle driver.

For each sample period t_N, the task manager 58 sends a request 94 to the GPS thread 62 to obtain CPS data. The task manager 58 uses a clock 96 as a reference for identifying each one second sample period. Each time a second passes according to clock 96, the task manager 58 sends out the request 94 that wakes up the GPS thread 62 to go read. the GPS data from the GPS satellite. Once the GPS data has been received, the GPS thread 62 passes the GPS data 96 to the sensor fusion thread 64. The GPS thread 62 then goes back into an idle mode until it receives another activation command from the task manager 58.

The task manager 58 can control when the CPS thread 62 is woken up. Instead of the GPS thread 62 being free running, the GPS thread 62 is operating according to a perceived time controlled by the task manager 58. The task manager 58 may send the activation request 94 to the GPS thread 62 once every second during normal sensor fusion operation. When the system is in a debug mode, however, the task manager 58 may only send one activation command 94. This allows the other operations performed by the system 89 to be monitored and determine how the single sampling of GPS data 96 propagates through system 89. The task manager 58 may also delay or disable task initiation to other threads, so that the processing of the GPS data 96 can be isolated.

The task manager 58 can isolate any state in the overall system 89, such as the state of system 89 after a first GPS reading by GPS thread 62 or the state of system 89 after the thirty second GPS reading by GPS thread 62 by controlling when and how often activation commands 94 are sent to GPS thread 62. In a similar manner, the task manager 58 can control when other tasks are performed by the system 89, such as when the sensor fusion thread 64 is activated.

Thus, the task manager 58 controls when Java applications are activated effectively running the overall system 89 in a lock-step mode. The task manager 58 can control the initiation of multiple tasks at the same time. This allows the task manager to control what parameters and operations are performed and used by the different Java threads so that different states in the multiprocessor system 89 can be detected and monitored more effectively.

One application for the task controlled applications is for accident reconstruction. The critical data manager 52 (FIG. 3) may save different vehicle parameters from a vehicle that has been in an accident. For example, sensor data, brake data, speed data, etc. The task manager 58 can feed the saved data into the different Java applications in a lock-step mode to determine how each Java thread processes the saved data. This can then be used to identify any failures that may have occurred in the system 89.

The system described above can use dedicated processor systems, micro controllers, programmable logic devices, or microprocessors that perform some or all of the communication operations. Some of the operations described above may be implemented in software and other operations may be implemented in hardware.

For the sake of convenience, the operations are described as various interconnected functional blocks or distinct software modules. This is not necessary, however, and there may be cases where these functional blocks or modules are equivalently aggregated into a single logic device, program or operation with unclear boundaries. In any event, the functional blocks and software modules or described features can be implemented by themselves, or in combination with other operations in either hardware or software.

Having described and illustrated the principles of the invention in a preferred embodiment thereof, it should be apparent that the invention may be modified in arrangement and detail without departing from such principles. Claim is made to all modifications and variation coming within the spirit and scope of the following claims.

Claims

1. A vehicle failure determination system comprising: modular processors linked over a network for operating different vehicle applications that control or monitor associated critical and non-critical vehicle operations;critical data associated with critical operations, wherein the critical data associated with at least one critical operation is related to at least one of sensor data, brake data, speed data, engine data and GPS data;an executive configured to operate in at least one of the processors, wherein the executive is configured to: identify and save into memory the critical data, wherein the critical data identified and stored into memory is saved in a lock-step mode to determine how each process thread was executed in the linked modular processors, and wherein the lock-step mode is determined by at least one of a system clock, a system watchdog timer, a message manager, a critical data manager, a security manager, and a data manager;operate a task manager to determine at least one of a potential system failure or system failure that has occurred in at least one critical operation and how each process thread saved into the memory in the lock-step mode identifies the failure.

2. The failure determination system of claim 1 wherein the network is at least one of wireless and wired links.

3. The failure determination system of claim 1 wherein the different vehicle applications include at least one of safety critical, mission critical and non-critical applications.

4. The failure determination system of claim 3 wherein the safety critical application includes at least one of driver assistance control, vehicle maintenance and accident reconstruction.

5. The failure determination system of claim 3 wherein the mission critical applications include at least one of car braking, car airbag deployment, car sensor monitoring, object detection operations, car audio processing, system diagnostics, GPS data, and driving operations.

6. The failure determination system of claim 1 wherein the sensor data includes at least one of infrared, radar, and video.

7. The failure determination system of claim 1 wherein the executive is secure and runs in a real-time operation, wherein the real-time operation pertains to applications in which the computer must respond as rapidly as required by the user or necessitated by the process being controlled.

8. The failure determination system of claim 1 wherein the task manager is used to reconstruct an accident.

9. The failure determination system of claim 1 wherein the memory is non-volatile.

10. A method of determining a system failure including: operating different vehicle applications on modular processors linked over a network that control or monitor associated critical and non-critical vehicle operations;monitoring critical data associated with critical operations, wherein the critical data associated with at least one critical operation is related to at least one of sensor data, brake data, speed data, engine data and GPS data;operating an executive configured in at least one of the linked processors, wherein the executive is configured to:identify and save into memory the critical data, wherein the critical data identified and stored into memory is saved in a lock-step mode to determine how each process thread was executed in the linked modular processors, and wherein the lock-step mode is determined by at least one of a system clock, a system watchdog timer, message manager, a critical data manager, a security manager and a data manager;operate a task manager to determine at least one of a potential system failure or system failure that has occurred in at least one critical operation and how each process thread saved into the memory in the lock-step mode identifies the failure.

11. The failure determination system of claim 10 wherein the network is one of wireless and wired links.

12. The failure determination system of claim 10 wherein the different vehicle applications include at least one of safety critical, mission critical and non-critical applications.

13. The failure determination system of claim 10 wherein the safety critical application includes at least one of driver assistance control, vehicle maintenance, and accident reconstruction.

14. The failure determination system of claim 10 wherein the sensor data includes at least one of infrared, radar, and video.

15. The failure determination system of claim 10 wherein the executive is secure and runs in a real-time operation, wherein the real-time operation pertains to applications in which the computer must respond as rapidly as required by the user or necessitated by the process being controlled.

16. The failure determination system of claim 10 wherein the task manager is used to reconstruct an accident.

17. The failure determination system of claim 10 wherein the memory is non-volatile.