TechSavvy Suite: iTech

TechSavvy Product Suite:
iSpec iTech iNspect (NEW!) iAudit (NEW!)

iTech
Smart troubleshooting. iTech is a model-based diagnostic engine which can be embedded in various applications ranging from structured guidance in helpdesk/ maintenance information systems to automated online fault isolation and remote diagnostics The iTech interactive troubleshooting assistant is used in applications such as e-Seaport's CODI-ExT CMDS (Crane Monitoring and Diagnostic System) a world first in automated, on-board real-time crane diagnosis. Documented benefits of using this troubleshooting technology are: Reduced downtime (Mean-Time-To-Diagnose reduced by up to 97%) Reduced training requirements Retention of specialist knowledge even with high staff turnover Quicker learning curve for maintenance staff Reduced instances of fault not found or incorrect diagnosis resulting in multiple parts replacement and higher costs
Knowledge capture/development The knowledge capture or development interface is a simple, easy to learn user interface allowing for the input of information on the system to be diagnosed. As it is browser based, it allows for developers anywhere on the internet to develop an application concurrently on a central server with only an internet or network connection. The approach used is called model based reasoning - where dependencies of the various system components are described using a tree type outline. Typically the system is broken down functionally (by subsystem) and then into components (spare parts, software, functional groups, etc.) For each component, all the relevant data can be input from a single screen: Component description Part information Test and repair information such as test times and procedures in text or pdf, XML, HTML, or any other windows applications Other interfacing data for automated testing/measurement functionality The user interface is simple and easy to use; it is also customizable to allow for application specific features and functionality.
Troubleshooting Once the data has been input, the system can be used to display information or to perform troubleshooting sessions. No fault trees or test patterns need to be generated. The system dynamically generates a test pattern using its built-in diagnostic logic. Unlike most other systems, the test pattern is optimized to reduce downtime and it does so using a unique set of algorithms which include probability theory, non-monotonic logic, inductive reasoning and Laplacian learning. So in brief (simplified description) when an alarm is triggered or a symptom reported the next-best-test to perform is always: The quickest test (downtime reduction) in the area with the highest probability of failure (probabilistic learning capability – start in the areas experience shows as the most likely causes) which will eliminate approximately half of the remaining suspect components (split half or binary search) The standard browser interface (completely customizable) will guide the user through the troubleshooting process (where not automated).