Defense spending opens door to system technology innovations

BY MIKE MACPHERSON, Vice President, Strategic Planning
Curtiss-Wright Defense Solutions
www.curtisswright.com

The U.S. Department of Defense’s (DoD’s) 2018 National Defense Strategy (NDS) said it clearly: “Our backlog of deferred readiness, procurement, and modernization requirements has grown in the last decade and a half and can no longer be ignored. We will make targeted, disciplined increases in personnel and platforms to meet key capability and capacity needs.” With that in mind, Congress increased the FY 2018 defense budget to $700 billion — an increase of $108 billion.

This article will lay out some of the areas where that budget will be spent and what areas may present opportunities for designers to innovate to close current and future technology gaps.

AI, big data, and robotics critical but need to be affordable
The technological priorities called out in the NDS will drive a significant increase in R&D spending to close technology gaps in advanced computing, artificial intelligence (AI), and autonomy and robotics. Among the priorities for modernizing key defense capabilities cited in the NDS that commercial off-the-shelf (COTS) vendors are well-positioned to support are:

• New investments in cyber-defense and the continued integration of cyber-capabilities into the full spectrum of military operations
• Investments in C4ISR to develop resilient, survivable, federated networks and information ecosystems
• Advanced autonomous systems, AI, and machine learning.

For developers of military embedded COTS electronics solutions, this additional spending promises increased support for technologies that address resilience, lethality, and readiness. Designers of defense and aerospace systems and platforms desire to continuously introduce advanced technology that provides the warfighter with an indisputable advantage in the battlefield. These technologies range from sensors, computing, and networking to electromechanical systems.

However, advanced technology by itself isn’t enough. It also needs to be affordable, reliable, and sustainable. The warfighters’ lives depend on the technology, and history has proven that if a soldier can’t trust their technology, they will abandon it.

New spending on advanced computing will result in improvements for leveraging big data analytics, enabling the warfighter immediate access to all of their critical information. Such access will require the use of cloud-computing technologies to enable data access by any device, wherever the soldier is located, at any time it’s desired. More than that, to bring the power of machine learning (ML) for AI to the network edge will require far greater local processing capability in order to deliver real-time data and solve the cloud’s inherent latency and bandwidth limitations.

Investments in AI and ML will provide capabilities that disrupt battlefield applications such as intelligence, surveillance, reconnaissance (ISR), and electronic warfare (EW). Supporting these new capabilities will require advances in heterogeneous high-performance embedded computing (HPEC) technologies.

Embedded systems for use on semi- and fully autonomous unmanned platforms, whether on the ground, in the air, or at sea, will require the development of low-power, ultra-small form-factor (USFF) processing, networking, full-motion video, and data-storage solutions. It’s estimated, for example, that a fully autonomous car will require 50 to 100 times the compute power needed to support today’s advanced driver-assistance systems.

The overarching investment strategy described in the DNS is to bring these advanced technologies to the battlefield in order to provide a force multiplier that gives warfighters a strategic and tactical advantage over the adversary. That said, it’s not enough to just deploy new technologies, it’s also necessary to ensure that those technologies are brought into the battlefield in a way that protects and secures them with the resiliency that they need to survive enemy attempts to disable or disrupt their intended operation.

Ensuring operational effectiveness in the field: GPS
The new technologies will provide new capabilities upon which the warfighter will surely become dependent. As such, they must also feature the defenses needed to ensure that their network and computing environments are protected against adversaries and so remain operationally effective.

An example of an advanced technology upon which the warfighter has become dependent is GPS. When introduced as part of the DoD’s Second Offset strategy in the mid-1970s, GPS provided a significant advantage in the battlefield thanks to its ability to deliver accurate position, navigation, and timing (PNT) data. This technology was essential for applications such as precision-guided weapons like the Tomahawk missile. Over the years, it’s become clear that our dependence on GPS also makes it a vulnerability. In environments in which GPS is denied or disabled, all of the weapons that depend on it are made ineffective. To counter that vulnerability and threat, an assured PNT (A-PNT) solution must be available that is able to operate even in a GPS-denied environment. New cost-effective and accurate COTS-based A-PNT technologies will enable the deployment of cost-effective, rugged solutions for GPS-denied environments.

Making AI and autonomous vehicles resilient
The development of new technologies based on AI will enable man-to-machine teaming solutions that deliver a significant advantage in the battlefield. Leveraging AI, autonomy, and robotics will result in machines that can operate independently, whether as an individual entity, paired with other machines in applications (such as a swarm configuration of drones), or in a soldier-machine interface in which the machine has its own autonomous capability augmenting the warfighter.

An example of the latter is an autonomous ground combat “mule” able to relieve the warfighter’s personal burden of carrying batteries, chargers, ammunition, etc. By reducing the weight in the warfighter’s backpack, these small autonomous vehicles will significantly increase the soldier’s ability to fight.

Likewise, the use of autonomous aerial vehicles to deliver logistics equipment or to locate IEDs will reduce the warfighter’s exposure to risk and improve their lethality. On the other hand, as these new solutions become common, adversaries will strive to find ways to attack and disable them. For example, one strategy for countering a learning machine is to spoof it with false information, forcing it to produce an incorrect answer.

Improving resilience, another key goal of the DNS, will ensure that deployed systems have the ruggedness and reliability to survive harsh environments and the security to protect against enemy attempts to exploit their vulnerabilities.

Autonomous vehicles, such as mine detectors, can keep the warfighter out of harm’s way, but that autonomy needs to be trusted. For this, the system requires the resilience, or self-resilience, that ensures that it’s reliable and can’t be easily disabled.

A machine can be manual, semi-autonomous, or fully autonomous. In each of these states, the higher the level of autonomy, the more the machine needs self-resilience. When a machine is fully manual, the warfighter provides the resilience. In the case of a semi-autonomous system, resilience is shared between the operator and the machine. In a fully autonomous system, resiliency depends completely on the expert systems built into that machine.

Autonomous systems need resilience and security
To be able to confidently depend on fully autonomous systems will require investments in technologies that provide both resilience and security.

An example of resiliency is found in safety-certifiable avionics systems for manned or unmanned military aircraft. To operate safely over domestic airspace, these platforms are increasingly required to meet DO-254 hardware and DO-178 software certification for specific Design Assurance Levels (DALs) recognized by aviation authorities around the world, such as the FAA in the U.S., the Canadian Transport Board, and EASA in Europe and the U.K. While safety certification is handled at the platform level, the electronic modules used to build out avionics subsystems must be supported with comprehensive data artifacts. Historically, modules for safety-certifiable subsystems were costly custom designs that took years to design and millions of dollars to develop.

In recent years, a new class of cost-effective DO-254-certifiable COTS boards has become available, greatly speeding and lowering the cost of integrating safety-certifiable applications. The preferred processor architecture for these COTS modules has been the Power Architecture family of devices being that Intel processors only support DO-254 up to the DAL C level.

As NXP shifts its focus from development of new Power Architecture processors toward Arm-based processors, designers of safety-certifiable systems are increasingly turning to Arm-based solutions. Arm processors support D0-254 up to the most stringent and critical level, DAL A, and also provide the additional benefit of very low power dissipation. The VPX3-1703 3U OpenVPX is a good example of an Arm-based single-board computer (SBC) (Fig. 1 ). It is designed for DO-254 safety-certifiable avionics applications.

Fig. 1: The VPX3-1703 is an example of an Arm-based 3U OpenVPX single-board computer designed for DO-254 safety-certifiable avionics applications.

The concepts of resilience and trusted systems refer not only to safety but also to data and hardware security. Great strides are being made today to enable COTS systems with anti-tamper technologies, cybersecurity, and protection of data-at-rest and data-in-motion.

For example, the Data Transport System (DTS1) network attached storage (NAS) device supports cost-effective two-layer encryption (Fig. 2 ). The DTS1 is also easily integrated into network-centric systems.

Fig. 2: Security in the field is critical, so the DTS1 NAS supports cost-effective two-layer encryption.

Design for tech-savvy warfighters
The soldiers now using this equipment are digital natives — almost born with modern technologies in their hands. Along with this technological adeptness comes a high level of assumption and expectation.

Today’s warfighter expects and depends on access to technologies as good as or better than what they have at home, such as an iPhone X, and social networking services to enable information sharing in real time in the battlefield. All of today’s internet resources, whether searching on Google or asking questions of Siri or Alexa, are only years away from being available to the warfighter. As we increasingly bring reliable networked desktop computing, mobile platform, and social media capabilities to the warfighter to enable “network-centric warfare,” the network itself has become a key component of our ability to operate.

This technological adeptness can also be leveraged to address readiness, an area of military spending that has been relatively underfunded in recent years. Advanced computing can be brought to bear for training and mission-planning and exploiting technologies developed for the gaming industry to provide sophisticated, realistic scenarios and experiences.

By having training embedded in the actual deployed platform, warfighters will be able to train while they operate without requiring a dedicated training location. Realistic simulation can be done virtually, providing, for example, the ability to train for a specific mission while en route.

Contain costs with open systems
Many of the technologies discussed above will benefit from the use of open systems, which reduce design risk and greatly speed time to deployment. The use of open systems also delivers significant cost reductions. Affordability results from competition and provides an alternative to expensive proprietary solutions.

Another key benefit of open systems is seen in technology insertions. Open systems enable the rapid insertion of new technology by defining an interface between different entities whose advancements progress at different rates. An open-systems interface, such as the OpenVPX system architecture, functions as a differential that enables the use of technologies that evolve out of synchrony.

For example, the fire control computer algorithms used in a main battle tank to handle ballistic solutions tend to evolve at a very slow relative rate with very little change from one year to the next. In comparison, the underlying processing technology used to run those algorithms progresses much faster. On the flip side, with EW as the example, the very sophisticated algorithms used to help identify a specific signal of interest in the noise of the electromagnetic spectrum have developed at a much faster rate than the processors that are used to run them in deployed systems.

The result is that the most advanced EW algorithms wait for processor bandwidths to catch up in order for them to be put to use. The use of open-standard interfaces enables the processing technology and the algorithms used on deployed platforms to advance at different rates.

Innovation opens door to vulnerabilities
For every new opportunity and technological leap forward, there is likely to be an associated vulnerability that emerges. While investing in the technologies sought by the DoD in order to enable new capabilities and increase force lethality, technology providers must also invest in mitigating against those vulnerabilities.

The use of COTS-based open systems provides a cost-effective approach to bringing these capabilities to the warfighter quickly and with the least risk. To bring the powerful benefits of advanced computing, AI, autonomy, and robotics to the warfighter, COTS solutions must be designed and packaged to meet the environmental and usage requirements of the battlefield. The equipment must be dependable and operate while exposed to extreme environmental conditions. The technology must also be designed and packaged to ensure safe and secure operation. Care must be taken to ensure safe operation without requiring burdensome safety precautions. System designers need to design and package next-generation COTS solutions to eliminate vulnerabilities to adversarial access or attack, including cybersecurity and protection against reverse-engineering to prevent physical access intended to disrupt operation.

It’s essential that these new technologies assure the security of the defense systems and critical information during development and operation.

Another area of great importance is testing, which must be done to ensure that deployed COTS solutions are reliable and deliver error-free operation throughout their useful life.

Conclusion
The DoD and warfighters depend on trusted and proven sources of supply, and Congress has made available the funds to make this happen. Now it’s up to designers and other innovators to realize the full promise of new technologies outlined here, just as examples. For sure, the COTS approach provides a proven alternative to costly, closed proprietary system architectures, speeds deployment, and ensures that critical technologies remain readily available over the lifecycle of their use. How technologists build upon and apply it for next-generation battlefield deployments with more tech-savvy warfighters will be interesting to watch.