Update - Data Centric Ecosystem Effort - Request for Information (RFI)

This is an update to the PEO STRI Data Centric Ecosystem Effort - Request for Information (RFI). Save the Date PEO STRI TableTop Exchange July 18 9:00 am - 12:00 pm EST PEO STRI, in collaboration with the Central Florida Tech Grove, is conducting a Tabletop Exchange for industry and academia to build up on recent market research responses on data centricity and Artificial Intelligence. Based on the outstanding industry feedback and collaboration, PEO STRI has narrowed the focus to the initial three problem areas with the most value to advance Army Transformation Initiative. Automation of training instructor, coach and analyst reports and other digital artifacts that include text, voice and video. Optimization of scenario generation with the goal to reduce significantly to plan and prepare lead time to training exercises at CTCs or Warfighter events. Automation of Level of Repair Analysis (LORA) to provide data-driven insights on repair levels, optimizing maintenance planning, enabling proactive product support solutions. We are seeking in your insights and ideas on how to address these challenges and help PEO STRI establish data pipelines for training and testing ranges into Army Enterprise Platform for interoperability, experimentation and decision-making activities. These efforts will be used to determine what follow-on initiatives PEO STRI might pursue to move toward a viable solution to fill this critical gap in soldier training performance analysis and acquisition streamlining. We invite you to join us to help shape our understanding of potential solutions to the problem described below. Plan to send your technical experts who can provide insights based on your organization's capabilities or that of other organizations that you know. Register to participate using the link below. During the registration process, you will be asked to select one or two of the solution capability areas. You will also be asked to upload a capabilities statement or digital brochure outlining your organization's relevant expertise. 1. What is the problem? STRI seeks to improve both the quality of the feedback that soldiers receive and decrease the amount of time that it takes to generate that feedback, in order that we can dramatically increase our efficiency in conducting training. To achieve this overall increase in efficiency, we need to automate the creation of the artifacts that provide both the instructors and trainees with the necessary summations of the training events and outcomes. Creating and planning for training takes a great deal of time and effort. In order to maximize our efficiency in providing training, we want to automate to the greatest extent possible how scenarios are generated and how simulation systems are prepared with the necessary particulars of the scenarios. Traditional LORA is typically a manual, labor-intensive, point-in-time analysis done during major design reviews. Once completed, it often remains unchanged even though mission conditions, failure rates, usage intensity, and costs evolve constantly. This makes product support strategies outdated and less cost-effective over time. 2. Why is it a problem? A trainer often has limited visibility of all the squad members and their actions and may not see the interaction between the soldiers. Therefore, the trainer has incomplete information as he tries to mentor the squad. The trainer may not have the knowledge or experience to coach each soldier, squad and above in corrective actions. Often a nine-member squad has two trainers resulting in an inability to adequately monitor and assess each soldier's performance. Currently, when a squad is part of a larger event, evaluating the squad's performance is a secondary goal. We need to evaluate all soldiers and units in a training event, regardless of the scale of the event. Continuous evaluation and feedback will increase the unit's performance. Currently, the trainer gives vocal feedback to the soldier and squad immediately after the squad runs a scenario. Sometimes this is sufficient. Other times, providing some type of visual feedback would enhance the soldier's understanding of how they can improve. Creating realistic and detailed scenarios can be a complicated and time-consuming process. We need to both reduce the time needed as well increase the simulation fidelity in order to maximize training efficiencies. It is both an issue of reducing complexities to make things simpler to implement as well as an issue of reducing the time needed to create the supporting artifacts, regardless of the form factor of the artifact (textual document, graphical diagrams, video presentations, web pages, etc.). The traditional LORA process relies on limited or stale data. The traditional process struggles with complexity and trade-offs. It is not responsive to operational changes. It is labor-intensive and costly to update. 3. What are some desired capabilities of a solution? Automation of training instructor, coach and analyst reports and other digital artifacts that include text, voice and video. There needs to be an AI capability either directly integrated into the PEO STRI training systems or - closely adjacent to the systems such that information can flow easily from the two systems. This AI system needs to collect data from the STE live, virtual, and constructive systems and blend it together as relevant for providing feedback to participants. Collection of data from the soldier's movements such as position (standing/prone), movement (walking/running), voice commands, and hand signals. Any device on the soldier must be small and light. Devices in the training area must account for trees and buildings. Collect of data from the soldier's weapon such as orientation, firing, and reloading. Any device on the weapon must be small and light. An AI agent that analyzes data from the soldier and squad and provides feedback on performance. A means of audio and visual feedback upon completion of the practice run at the training site. The equipment must be small and light, easy to use, and use local power (battery, vehicle, etc.) Optimization of scenario generation with the goal to reduce significantly to plan and prepare lead time to training exercises at CTCs or Warfighter events. Regarding creating scenarios and populating the systems with the necessary elements, an AI system should be designed such that it proactively engages with a training exercise/simulation designer to solicit all the elements needed for the scenario. In short, the AI system should dialog/interview the designer to determine the bounds of the exercise and any specific requirements that are needed, and then the AI system should create from that. An AI system should prepare elements for an exercise that provide the desired/sufficient level of realism and complexity for the virtual component of the exercise and build/use the virtual elements as needed. An AI enabled application should plan for the appropriate blending of live, virtual, and constructive elements into the scenario and produce easy to follow descriptions for the exercise planners about what is being provided in each environment. And whereas the AI system could prepare the scenarios in the virtual and constructive systems, the AI system would need to give instructions on how the live training areas would need to be configured (if deviating from an established physical setup in the live training area). An AI system would be able to store exercise plans that it had generated with a designer such that they could be retrieved and augmented/updated at a later date. Optimize the LORA process for PEO STRI training systems and capabilities allowing product support managers and logistics management specialists the ability to allocate resources appropriately to lead to more effective and efficient maintenance strategies. Predictive analytics engine: Forecast failure modes, usage patterns, and repair demand using machine learning. Dynamic cost modeling: Automate updates to cost elements (labor, spares, maintenance) based on real-time logistics data. Scenario simulation: Run multiple LORA scenarios to compare cost/benefit for field and sustainment/depot-level repairs under different operational conditions. User-configurable policies: Allow input constraints (e.g., workforce limitations, facility availability, mission criticality). Optimization algorithms: Use AI to recommend the most cost-effective repair policy mix, balancing total ownership cost and readiness. Visualization dashboard: Provide clear, intuitive outputs showing cost trade-offs, break-even points, and repair timelines.