Curtiss-Wright Defense Solutions

The future of combat will increasingly leverage both  unmanned and optionally unmanned fighting vehicles.  A key driver for the use of unmanned platforms is their  ability to collect data without putting warfighters in  harm’s way. A resulting challenge is how to store the massive  amounts of sensor data collected onboard the unmanned  platform and how to ensure that data is secure in case the  platform should get lost and fall into the wrong hands.Adding  to this problem is the inexorable rate of increase in sensor data  resolution. For example, where we had HD video just a few  years ago, we are now seeing requirements to support 4K  video, which will itself soon be supplanted by 8K video.

To turn all of this valuable data into actionable  intelligence requires significant amounts of processing, some  of which can take place onboard the platform if powerful and  rugged enough compute resources are available. Onboard  processing enables a reduction in the size of the data,  enabling key data to be downloaded in realtime to analysts at  the Forward Operating Base (FOB). Unfortunately, the data  downlink transports available from unmanned vehicle shave  been unable to keep pace with the firehose of data that these  platforms are now able to collect and store.

The good news is that advances in processing  technologies, such as the use of GPU enabled devices to  drive AI and ML applications, can help optimize data sizes  to fit through a realtime pipe. Using a “store and forward”  approach, a rugged high density storage system onboard the  platform can be used to store and protect all of the collected  data at full resolution for post-mission analysis, after the  drone, for example, returns to base. During the mission,  subsets of sensor data (think of low resolution thumbnail  images) can be created by compressing the data or adjusting  the sampling rate to produce an acceptable and “good enough”  representation of that data to fit into the realtime transport pipe  for immediate transmission. This approach can quickly provide  usable information, such as sensor, video, positional, thermal,  or fuel data, for example.What’s more, forwarding thin-pipe  level data in realtime speeds the process of identifying what  data is most important for analyzing once the full resolution  video and data stored on the platform returns to the FOB,  where it can be reviewed post-mission, similar to how an NFL  football coach can study in detail the recording of Sunday’s  game on Monday.

During some unmanned missions, probability of intercept/  probability of detection concerns will make the realtime  downlinking of data untenable. Once  a vehicle is sent out into the battlefield  there’s a greater chance of the vehicle  and the critical data it has collected  falling into adversarial hands. For  those reasons the sensor data must be  stored using a system with Data-at-  Rest (DAR) protection. By leveraging  multi-layer commercial encryption,as  outlined by the NSA’s Commercial  Solutions for Classified (CSfC)  program, critical data can be encrypted  for storage and forwarding in an NSA  approved manner. This provides an  alternative to Type 1 encryption with  greatly reduced risk and consequences  in the event the platform or data is  captured by the enemy. A good example  of a data storage system that can protect  terabytes of DAR using CSfC is Curtiss-Wright’s DTS1  Network Attached Storage solution.

 In the case of potential  data breaches,  there is no magic  pill but continuous  improvement through  awareness training,  suitable technologies  and occasional testing  will help to safe-guard  organisations 

Unlike airborne platforms, unmanned ground vehicles  and naval platforms can also benefit from recent advances  in data communication. For example, the advent of 5G  networks in the battlefield can establish a large high-speed  data communications “bubble” that enables collected data to  be processed in the field in realtime.

Typically, the size of the vehicle and its distance from the  FOB will determine whether or not sensor data transport can  be performed in realtime. An example of a rugged processing  system for data analysis is Curtiss-Wright’s PacStar Tactical  Fusion system, which can be used to search for signal of  interest in data captured from drones in realtime to the FOB,  or can be used to process the full high resolution data set  after removal from the returned unmanned platform. Today,  sophisticated battlefield networks can be quickly stood up to  connect the FOB to the Cloud. What’s more,if connectivity is  denied or reduced, the Cloud can be replicated locally at the  FOB for limited amounts of times. Curtiss-Wright’s PacStar  MDC system is a compact, mobile system that can act as a local  Cloud or as a Cloud synchronization point, so that multiple  battalions in theatre can share a disconnected Cloud. This  enables unmanned vehicles to continue transmitting sensor  data, which is stored temporarily in the replicated Cloud, and  when connection is restored, can be re-synchronized with  the appropriate government Cloud, making the critical data  available to leadership.

If there’s one biggest technological advance making  possible the recent breakthroughs in the miniaturization and  ruggedization of deployable enterprise  networking/computing solutions  it’s the advance of modern GPU  hardware, the software that leverages  those GPUs, and the availability of  ruggedized versions of those devices,  from vendors such as NVIDIA, that  can be used in deployable military  systems. Additional factor helping  to drive these capabilities are the  increase in solid state storage density  that’s appropriate for field usage, such  as NVMe devices, and the inevitable  march of Moore’s Law that continues  to drive new levels of processing  power on a regular drumbeat.

Along with the ability to collect  and store sensor data in the field, the  DoD’s mandate to use electronics  systems based on the modular open system approach (MOSA)  has helped increase the use of standardized interfaces such as  Ethernet and IP networks. Where before system designers may  have selected, for example, dedicated video buses like SDI,  that were difficult to integrate, the use of Ethernet is becoming  ubiquitous. Previously, a platform might have deployed an  analog sensor that required a dedicated piece of translation  equipment to convert to the data into digital format. Today,  advances in electronics enables those sensors today to directly  use a common data format or use a published API. That means  that the unstructured store of data, sometimes referred to as  a “data lake,” can be more rapidly exploited using powerful  analytic software, such as Kinetica, which can process the  unstructured data at a rate not even imagined a few years ago.  All of these advances combine to empower the warfighter  and decision makers with more data, which helps them observe,  orient, decide, and act on the most informed path.