Technology is the central pillar of modern warfare. Paul Maxwell, Lt. Col. (Retd.) of the US army said the use of AI would improve capability and resilience, and the military cannot afford to ignore this: “Certainly our opponents will and we must be able to attack and defeat their AIs,” he said. In the light of the ever-changing geopolitical scenario, Joint Artificial Intelligence Center (JAIC) – AI Centre of Excellence under Department of Defense (DoD), US has come out with an action plan to facilitate the integration of AI across departments, especially the Department of Defence (DoD) by 2025.
Pentagon’s AI artillery
The Pentagon is counting on AI as a major driver of defence modernisation. The US government has allocated the Pentagon around $841 million for building AI capabilities for 2021, of which more than one-third will go to the JAIC. The DoD has launched a program, called Tradewind, incorporating tech players, academia, and other stakeholders to develop new AI capabilities for the defence forces. Additionally, the Pentagon has started a $10 billion program to centralise data across departments over the cloud.
“Tradewind will provide a user-friendly framework for our private sector partners to work more efficiently with the DoD to scale and implement AI for the warfighter and consumers across the military. We want to learn from this initiative to improve the way DoD works with all types of private sector and academic partners, and inject the much-needed speed and agility necessary to scale artificial intelligence and transform the Department,” said William Roberts, JAIC Chief of Acquisition.
The Tradewind portal will be available via AI.mil, the Department of Defense’s main artificial intelligence website. The first step of the project will be to develop online capabilities that communicate with small and non-traditional businesses, create rapid contract actions, and assist DoD offices with end-user integration through agile delivery. The Data Readiness for Artificial Intelligence Development (DRAID) will allow defence organisations to issue orders for data preparation work, which may include anything from data collection to data sorting and storage to modelling how workers can use AI to gain deeper insights.
What’s India upto
While speaking at the e-symposium event organised by SAMDeS along with the FICCI, RKS Bhaduria, IAF Chief Marshal, expressed IAF’s interest towards integration of the AI concepts for future combat operations. In a recent news article, the Ministry of Defence (MoD) mentioned the utilisation of low latency communication provided by the 5G technology to power up AI and build UAV for the armed forces. The MoD has established iDEX – an ecosystem to promote defence and aerospace innovation and technology growth by bringing together innovators and entrepreneurs to provide technologically innovative solutions for the modernisation of the Indian military. Take a look at the recent winner for remote real-time in-flight health monitoring of aircrew:
Moreover, it has set up a Technology Development Fund (TDF) under the ‘Make in India’ program. The program is executed by DRDO to encourage industries participation in cutting edge technologies for defence applications. Feasibility studies are under way for various defence projects. DRDO is trying to develop SSPA (Solid State Power Amplifier) – a component of transmitter system, for AMDR 2D/3D radar for the Indian Navy. For Indian Air Force, DRDO is in the process of building a wireless aircraft flight data recorder and automatic weather station.
India’s IITs, IISc, NITs, and IISERs are home to remarkable academics in computer science and engineering. To integrate AI capabilities in defence, a convergence between academia, industry, and policy is critical.
AI is a game of collaboration. Good datasets are crucial to building good AI models. But building models for the military is even trickier. The nature of data getting collected, the risk of mask surveillance in the name of counter insurgency, and an absence of a strong data protection bill make data collection problematic. So it’s important to have an airtight policy in place on the use of public data for building military AI capabilities.
That said, the military could enlist private players and co-opt their AI/ML models to build defence capabilities. AI works best in collaboration, not in siloes. But when we use AI, we should have the guardrails – like anonymised and unbiased data –to ensure fairness.