Introduction to Cervus • We were formed in 2013 • We come from Force Development and Collective Training Backgrounds • We exploit military and security training data to provide our customers with a comprehensive understanding of their operational performance. • Using pioneering data capture systems, industry standard analytics and secure data storage solutions, we provide services to exploit your training data.
The Collective Training System MANAGEMENT RULES • Environment System – what is CAPTURE SYSTEM required to create the appropriate SCENARIO Environment ENVIRONMENT • Scenario System – what is EXPLOITATION required to wrap around the SYSTEM environment with a logical and consistent scenario • Capture System – what is required to capture appropriate contextual data and performance data on the training subjects • Analysis System – that which is ANALYSIS SYSTEM required to collate, analyse and manage data • Exploitation System – what is required to exploit the data for AAR and as long term actionable insights
The central idea is unlocking power through applying a concept of exploitation [of] what we can learn from current operations, from our training and from experimentation within training. Commander Force Development and Training (Comd FDT) July 2011
Balancing Act Requiring Known costs Limited opportunity for Unknown costs and and budgeted repeatability repeatability limited budget Focussed on the Data capture Focussed on the Data capture to achievement of typically to experimentation inform measures of Collective Training support an after objectives and performance / Objectives (CTOs) action review evidence effectiveness Training Experimentation
Our Vision for Training Data Exploitation Harvest LVC and Objective Intelligent Analytics Broad All Data C2 Data Analysis Utility
From selective data gathering to harvesting all data . • The reduced cost, size and power consumption of modern sensors, power storage technologies, communications systems and data storage devices means there is no longer a need to limit data gathering to small data sets. • It is now technically possible to harvest and record virtually all data, linked to all participants (BLUEFOR, OPFOR, wider COEFOR and observers themselves) associated with every training event: allowing for an exponentially richer training data set to mine and exploit – enabled by a “capture everything now, use anything later” approach.
From use of Live training data to the use of Live, Virtual, Constructive & Command data . • Traditionally, the main focus for training data gathering has been within the Live training environment, and then mostly positional and firing data. However, there is a plethora of wider useful training data that can be harvested: human biometrics; vehicle usage, performance and health; C4I (voice, data and meta-data); ISR; topographical and meteorological data to name but a few. • Also as the Live, Virtual and Constructive training environments continue to blend ever more seamlessly into a single synthetic training environment, there is an opportunity to draw (and merge) training data from each of these synthetic domains to provide a far more holistic training data set to mine and exploit. • Comms data and HUMS are both undervalued and underused training data sources.
From primarily gathering subjective data to the harvesting of objective data . • Whilst some limited objective data is currently gathered, new and emerging technologies allow for the routine and automatic harvesting and storage of far more, and far wider, objective data sets. • This will free the training observer to concentrate more on expertise-based subjective observation but, as importantly, will provide a wealth of context within which to ultimately frame far more meaningful and useful insights from training.
From manual analysis to automated, intelligent (AI) analysis . • The relentless development of ever-more ‘intelligent’ machines provides an exceptional opportunity to move rapidly away from the resource-heavy, time-consuming activity of manual data analysis to an automated – even intelligent – analytical approach. • Self-adapting algorithms, pattern recognition technologies and machine learning approaches now mean that the drawing of meaningful insights from masses of data is simple, time-efficient, ever-improving, self-teaching and increasingly affordable. The obstacles that the manual processing and analysis of large amounts of training data once presented are now easily surmountable.
From descriptive analysis to predictive & prescriptive analysis. • This machine learning capability now allows for a genuine step-change from purely Descriptive analytics (What has just happened? What is happening now?) to Predictive analytics (What is likely to happen next, based upon experience?) to Prescriptive analytics (What could/should be done about what is likely to happen next, so as to achieve a positive outcome?). • From retrospective after-action consideration of training data to real-time interpretation and understanding . • From observed facts (pure data) to informed insights (via analysed data in context). • From simple observation of training, mostly after the event, to helpful, proactive interventions during (and even before) training.
From limited to broad utility . • By making use of open architectures and common standards and modern cloud-based storage and processing technologies, the training data gathered, and the analysis drawn from it, will be of use not just to the immediate training community but also to individuals and the wider field army , force development, research, experimentation & acquisition communities – all of whom will be able to access the data they need whenever, and from wherever, necessary.
HIVE- a solution • HIVE will be unparalleled in its adaptive ability to harvest, categorise, store, and analyse data from collective training environments. • More than that, HIVE can deliver genuinely comprehensive and exploitable insights via its innovative machine learning engine and unique visual reporting systems : thus, allowing commanders to quickly spot and interpret trends from training, gain context-derived insights from them, and to rapidly and clearly identify opportunities for enhancements to warfighting. • HIVE can truly build ‘winning foundations’ .
You need judgement and creativity to determine how to find solutions to what the data is telling you, but those judgements, in turn, are tested as part of the next optimisation loop. Creativity not guided by a feedback mechanism is little more than white noise. Success is a complex interplay between creativity and measurement, the two operating together, the two sides of the optimisation loop. ” Matthew Syed
HIVE Demonstration Collective Training Event
Assessment Programme Id Name Start Date End Date Id Msn 4 Events 1 Battle Preparation Initial 2 Vehicle Move to LoD 1 Assessment 07/05/2018 07:35 07/05/2018 12:13 2 Mission 1 07/05/2018 13:35 07/05/2018 16:13 3 Pre designated Fires on Enemy Target 3 Mission 2 08/05/2018 07:35 08/05/2018 10:11 4 Adjustment of Fires 4 Mission 3 08/05/2018 12:35 07/05/2018 14:10 5 Fire Support from MIV 5 Mission 4 09/05/2018 07:35 09/05/2018 12:26 6 Cross LoD 7 Assault Building 1 Room 1 Initial 8 Assault Building 1 Room 2 6 Assessment 08/05/2018 06:55 08/05/2018 12:13 9 Assault Building 1 Room 3 7 Mission 1 08/05/2018 12:35 07/05/2018 14:10 8 Mission 2 09/05/2018 07:35 09/05/2018 12:26 10 Enemy Fires onto BLUEFOR 9 Mission 3 09/05/2018 14:13 09/05/2018 16:16 Enemy Counter Attack from Building 2 onto 11 Building 1 10Mission 4 10/05/2018 07:05 10/05/2018 14:15 Exploit and Assault from Building 1 to Initial 12 Building 2 11 Assessment 09/05/2018 07:05 09/05/2018 12:11 13 Assault Building 2 Room 1 12Mission 1 09/05/2018 14:40 09/05/2018 16:43 14 Reorganisation 13Mission 2 10/05/2018 07:05 10/05/2018 12:11 15 Casualty Extraction 14Mission 3 10/05/2018 14:10 10/05/2018 16:32 16 POW Extraction 15Mission 4 11/05/2018 07:09 11/05/2018 12:23 17 Hot Washup 18 Training System Reset
Scenario and Environment Simulation: Fires Bde Fires • HICON Coy HQ Noise Fires BG Fires Observer Mentor Primary OM • Smell Observer Mentor Secondary OM Observer Mentor Primary OM Observer Mentor Secondary OM • Blast /Vibration Dismount Sect Comd Dismount Sharpshooter Dismount Grenadier Enemy Dismount Commander Dismount Grenadier Enemy Dismount Sharpshooter Dismount NLAW Enemy Dismount Grenadier Dismount ASM Enemy Dismount Anti Tank Dismount Sect 2IC Dismount Rifleman MIV AFV Crew Driver Crew Gunner Crew Commander Enemy Fires Combat Team Fires Building Enemy Position Building Enemy Position
Capture Systems • • I-DES Warrior Metric • I-DES • • Joint Fires Virtual Training • Joint Fires Virtual Management and • AWES Capture Tablets • AWES • I-DES Current Systems • I-DES • I-DES • DFWES • DFWES • AWES • AWES • Joint Fires Virtual • HR Monitor • HUMS • DFWES • Joint Fires Virtual • AWES
Analysis and Exploitation
Analytics Objective Intelligent Data Analysis LVC and C2 Harvest Broad All Data Utility
Recommend
More recommend
