Mr. Gersh will review relevant work at JHU/APL and discuss two projects in particular

  

Mr. John Gersh

 

Intelligence analysts face the daunting problem of navigating large amounts of information to build a story within a domain of analysis. Some of this information is often represented as a semantic network, a set of entities and the relationships that connect them, all richly typed and characterized. In today’s operational environment analysts must also respond to a constant flow of information with results that can be understood and acted upon rapidly. We are pursuing an integrated approach to this challenge. We are investigating methods to support the ongoing process of analytical sensemaking and to support analysts’ sharing of insights with colleagues, customers, trainees, and their future selves. Together with colleagues from Klein Associates, we have built a prototype system based on a conceptual model of sensemaking in intelligence analysis. The system provides capabilities for visualizing, manipulating, and storing insights as dynamic collections of information together with visual representations of concepts that explain their analytical roles.

Autonomous behavior is an increasingly important component of many engineering systems, ranging from fault-tolerant spacecraft and unmanned autonomous vehicles to non-player actors in computer games. Autonomous systems are particularly challenging for human-stem interaction: they are reactive, operating in real time in response to specific input events, and are supervised, rather than controlled by human users. One approach to dealing with this challenge is through the use of visual techniques. This project used the NASA Solar-Terrestrial Relations Observatory (STEREO) spacecraft, built and operated by JHU/APL, as the example system for building a research prototype. A set of capabilities of this prototype were expanded and integrated into the STEREO ground system with mission funding; this system has been in operational use in the STEREO mission for some time. JHU/APL is now developing a system for spacecraft autonomy design and supervision using finite state machines that extends these capabilities into a new domain.


John Gersh is a member of the Principal Professional Staff of The Johns Hopkins University Applied Physics Laboratory, in the Laboratory’s System and Information Sciences group. His interests involve investigating and implementing the interactive depiction of an information-rich environment to decision makers and developing knowledge representations and software architectures to enable those interactions. He has also worked extensively in the area of human interaction with complex automation and autonomous systems. During his twenty-eight years at JHU/APL he has participated in system development and led research teams in these areas in domains as diverse as Navy shipboard and Battle Group systems, airport operations, spacecraft autonomy, and intelligence analysis. He has worked extensively in the intelligence community in research and development of new information technology and its integration into complex information systems. He has also worked on NASA technology-development programs, recently going from a concept prototype to an operational mission ground system. He was also instrumental in the establishment of a Laboratory-wide R&D program in cognitive engineering at JHU/APL. Mr. Gersh received S.B., S.M., and E. E. degrees in electrical engineering from M.I.T., and also studied philosophy at Harvard, which has turned out to be surprisingly useful in thinking about people interacting with information and with the world.