Some projects are simply too big to finish on the release schedule that you want to maintain. Ral... more Some projects are simply too big to finish on the release schedule that you want to maintain. Rally's agile engineering team delivered a very disruptive architectural change lasting nine months in parallel with three releases. This paper will show how this was done without prolonging the usual release cycles and without technically crippling our product or doing a lot of throwaway work. The main elements of our approach were: the "Sacrifice One" pattern, incremental delivery of the architectural change when possible, technical infrastructure modifications to support incremental changes to the architecture, and addressing the biggest risk first. When we were close to finished, we added the whole team for the final push to release with the new architecture. The end result was three successful releases of the existing system prior to the final release rolling out the new architecture.
We present an early vision of a system that allows computer controlled rendering of physical surf... more We present an early vision of a system that allows computer controlled rendering of physical surfaces, terrains, and environments by manipulating grids of "physical pixels" or rods whose heights can be raised and lowered on command. A user would be free to walk within such a dynamically deformable physically rendered environment (PRE). The system would be able to create on demand the floor, ceiling and sides of the "Holodeck" in which the person resides, and vary the slipperiness of the surfaces to provide authentic tactile feedback in real time. Ideally, the system would support discreet relocation while a person walked to create the impression of infinite distance. Such a system could be combined with immersive graphic visualization and futuristic programmable matter to convey realistic impressions of a walking tour of a remote site, the experience of climbing a steep mountain, or navigation through a maze. The potential applications of this system range from entertainment to police and infantry training to education, and extend as far as the imagination allows. We propose mechanisms with which we believe that such a vision could be implemented with the technology of today. This paper is an early work intended to raise interest in this idea, as opposed to presenting a finished solution for achieving the vision of a Holodeck.
We present an early vision of a cyber-physical environment in which computer controlled rendering... more We present an early vision of a cyber-physical environment in which computer controlled rendering of physical surfaces, terrains, and environments is achieved by manipulating grids of "moving physical pixels" or "moxels", whose heights can be raised and lowered on command. A user would be free to walk within such a dynamically deformable physically rendered environment (PRE). The system would be able to create on demand the floor, ceiling and sides of the "Holodeck" in which the person resides, and vary the slipperiness of the surfaces to provide authentic tactile feedback in real time. A user would be able to conduct a walking tour of a remote site, the experience of climbing a steep mountain, or navigation through a maze -all within the bounds of the Holodeck. Multiple users could be networked together to create a distributed cyber-physical system to enable team-based distributed training. We present early results in a Holodecklike simulator, HoloSim.
Software failures in wireless sensor systems are notoriously difficult to debug. Resource constra... more Software failures in wireless sensor systems are notoriously difficult to debug. Resource constraints in wireless deployments substantially restrict visibility into the root causes of node-level system and application faults. At the same time, the high costs of deployment of wireless sensor systems often far exceed the cumulative costs of all other sensor hardware, so that software failures that completely disable a node are prohibitively expensive to repair in real world applications, e.g. by on-site visits to replace or reset nodes. We describe NodeMD, a deployment management system that successfully implements lightweight run-time detection, logging, and notification of software faults on wireless mote-class devices. NodeMD introduces a debug mode that catches a failure before it completely disables a node and drops the node into a stable state that enables further diagnosis and correction, thus avoiding on-site redeployment. We analyze the performance of NodeMD on a real world application of wireless sensor systems.
Software failures in wireless sensor systems are notoriously difficult to debug. Resource constra... more Software failures in wireless sensor systems are notoriously difficult to debug. Resource constraints in wireless deployments substantially restrict visibility into the root causes of node-level system and application faults. At the same time, the high costs of deployment of wireless sensor systems often far exceed the cumulative costs of all other sensor hardware, so that software failures that completely disable a node are prohibitively expensive to repair in real world applications, e.g. by on-site visits to replace or reset nodes. We describe NodeMD, a deployment management system that successfully implements lightweight run-time detection, logging, and notification of software faults on wireless mote-class devices. NodeMD introduces a debug mode that catches a failure before it completely disables a node and drops the node into a stable state that enables further diagnosis and correction, thus avoiding on-site redeployment. We analyze the performance of NodeMD on a real world application of wireless sensor systems.
Some projects are simply too big to finish on the release schedule that you want to maintain. Ral... more Some projects are simply too big to finish on the release schedule that you want to maintain. Rally's agile engineering team delivered a very disruptive architectural change lasting nine months in parallel with three releases. This paper will show how this was done without prolonging the usual release cycles and without technically crippling our product or doing a lot of throwaway work. The main elements of our approach were: the "Sacrifice One" pattern, incremental delivery of the architectural change when possible, technical infrastructure modifications to support incremental changes to the architecture, and addressing the biggest risk first. When we were close to finished, we added the whole team for the final push to release with the new architecture. The end result was three successful releases of the existing system prior to the final release rolling out the new architecture.
We present an early vision of a system that allows computer controlled rendering of physical surf... more We present an early vision of a system that allows computer controlled rendering of physical surfaces, terrains, and environments by manipulating grids of "physical pixels" or rods whose heights can be raised and lowered on command. A user would be free to walk within such a dynamically deformable physically rendered environment (PRE). The system would be able to create on demand the floor, ceiling and sides of the "Holodeck" in which the person resides, and vary the slipperiness of the surfaces to provide authentic tactile feedback in real time. Ideally, the system would support discreet relocation while a person walked to create the impression of infinite distance. Such a system could be combined with immersive graphic visualization and futuristic programmable matter to convey realistic impressions of a walking tour of a remote site, the experience of climbing a steep mountain, or navigation through a maze. The potential applications of this system range from entertainment to police and infantry training to education, and extend as far as the imagination allows. We propose mechanisms with which we believe that such a vision could be implemented with the technology of today. This paper is an early work intended to raise interest in this idea, as opposed to presenting a finished solution for achieving the vision of a Holodeck.
We present an early vision of a cyber-physical environment in which computer controlled rendering... more We present an early vision of a cyber-physical environment in which computer controlled rendering of physical surfaces, terrains, and environments is achieved by manipulating grids of "moving physical pixels" or "moxels", whose heights can be raised and lowered on command. A user would be free to walk within such a dynamically deformable physically rendered environment (PRE). The system would be able to create on demand the floor, ceiling and sides of the "Holodeck" in which the person resides, and vary the slipperiness of the surfaces to provide authentic tactile feedback in real time. A user would be able to conduct a walking tour of a remote site, the experience of climbing a steep mountain, or navigation through a maze -all within the bounds of the Holodeck. Multiple users could be networked together to create a distributed cyber-physical system to enable team-based distributed training. We present early results in a Holodecklike simulator, HoloSim.
Software failures in wireless sensor systems are notoriously difficult to debug. Resource constra... more Software failures in wireless sensor systems are notoriously difficult to debug. Resource constraints in wireless deployments substantially restrict visibility into the root causes of node-level system and application faults. At the same time, the high costs of deployment of wireless sensor systems often far exceed the cumulative costs of all other sensor hardware, so that software failures that completely disable a node are prohibitively expensive to repair in real world applications, e.g. by on-site visits to replace or reset nodes. We describe NodeMD, a deployment management system that successfully implements lightweight run-time detection, logging, and notification of software faults on wireless mote-class devices. NodeMD introduces a debug mode that catches a failure before it completely disables a node and drops the node into a stable state that enables further diagnosis and correction, thus avoiding on-site redeployment. We analyze the performance of NodeMD on a real world application of wireless sensor systems.
Software failures in wireless sensor systems are notoriously difficult to debug. Resource constra... more Software failures in wireless sensor systems are notoriously difficult to debug. Resource constraints in wireless deployments substantially restrict visibility into the root causes of node-level system and application faults. At the same time, the high costs of deployment of wireless sensor systems often far exceed the cumulative costs of all other sensor hardware, so that software failures that completely disable a node are prohibitively expensive to repair in real world applications, e.g. by on-site visits to replace or reset nodes. We describe NodeMD, a deployment management system that successfully implements lightweight run-time detection, logging, and notification of software faults on wireless mote-class devices. NodeMD introduces a debug mode that catches a failure before it completely disables a node and drops the node into a stable state that enables further diagnosis and correction, thus avoiding on-site redeployment. We analyze the performance of NodeMD on a real world application of wireless sensor systems.
Uploads
Papers by Veljko Krunic