Computer Vision (CV) has direct applications for satellite and drone use; both create large amounts of data which takes human analysts significant time to examine to search for relevant information. CV has the capacity to comb through this data significantly quicker and decrease the amount of time it takes to obtain actionable information. CV also has direct applications in military technology and its future utilisation will include automated military vehicles as well as missile targeting, and missile defense technology.

Listed below are the key areas in which CV is impacting the aerospace and defense industry, as identified by GlobalData.

Autonomous land vehicles

GlobalData predicts that, between 2020 and 2025, the first fully autonomous vehicles are expected to be operating in limited zones of certain cities. This prediction demonstrates that civilian autonomous vehicles are increasingly advanced. In the military sector, there is an increasing push for unmanned land vehicles, either armed or unarmed. In the US, the Dod is currently seeking bids for the ATLAS program, an advanced automated program which will is designed to potentially be able to control tanks in the future, though with a human operator in the loop.

Fully autonomous drones

Future drones will utilise CV for two reasons, firstly for visual analysis – drones can capture large amounts of visual data using high definition cameras and this currently requires significant work by human operators for analysis. In the future, CV will be used to quickly examine visual data and extract important information. The second utility of CV will be for increased autonomy; currently, drones require a human operator but in future will be almost completely autonomous, potentially gaining the ability to detect and engage targets.

CV for satellite data

Satellites create large amounts of actionable data which CV can help process. This has applications in both the commercial and military sectors. In the military sector, CV could be utilised to locate military architecture or track weapons systems in real-time.

Targeting systems

In line with the projected use of CV with satellite and drones, CV in this context can be utilised to enhance or automate targeting systems, this will be an essential development with faster missile systems which require reactions faster than a human operator can provide. CV does not have to be used to fully automate weapons targeting systems but can be used to increase the accuracy of existing weapons systems. These targeting systems are expected to be applied to future hypersonic weapons, as well as to new missile defence systems.

Tech controversies

A barrier that states will have to overcome is how to integrate tech companies into defence programs whilst overcoming ethical controversies. Project Maven encountered significant difficulties because Google withdrew following opposition from its staff. There is a requirement at present for nondefense companies to be involved with these advanced projects because of their specialist knowledge, and the ethical standards of tech companies will have to be addressed moving forward.

Homeland security

CV technology has significant homeland security applications; the technology can be utilised to protect facilities and to identify people, as well as search for weapons and vehicles. Contracts have already been issued for uses of this technology; Trueface will be providing applied computer vision at air force bases in the US. This technology can scan crowds to search for individuals or to locate and track cars across cities.

Arms manufacturing

CV has utility for manufacturing, including arms manufacturing. The first is for monitoring complex manufacturing by human operators, in the case of a manufacturing process where an operator has to make many steps; CV can ensure that no steps are missed. The second is quality control; CV can scan outgoing products, munitions, and remove any damaged products from the supply chain.

This is an edited extract from the Computer Vision in Aerospace and Defense – Thematic Research report produced by GlobalData Thematic Research.