As world in general begins to move towards widespread use of artificial intelligence (AI for short) authentic data becomes increasingly important. Battlefields of the future are no exception. But for the AI to be effective in any field, the necessary input data is essential. It is often said that ‘data is the new oil’. But that data has to be authentic and genuine or else the adage in case of computers ‘garbage in-garbage out’, may prove to be true.
Armed forces, especially the traditional ones like the Indian one, have been slow to adapt to new technological environment precisely due to neglect of data gathering. In my several years of research in military history, I have found some amount of statistical data but very little operational data. Let me explain- the various formations involved in counterinsurgency operations or cross border small operations indeed periodically give out the numbers. The numbers generally deal with casualties.
There are two other ways that operational data is captured presently. One method is the after action reports in case of major operations. These are prepared by the very unit/formation that undertook the operation. It is hardly likely that these can contain any critical inputs or mistakes/miss-steps. Without much ado these can be called a white washing job. Secondly, when there is a tactical disaster or a failed operation, a court of inquiry is ordered. The proceedings of court of inquiry (COI) are indeed independent and do generate authentic data. But the focus of COI is legal and not tactical and the aim is to find violations of law/procedures etc and to punish the guilty. For instance the technical details like performance of weapons etc are never part or the aim of such proceedings.
In the current situation we do not have an independent organisation to quickly visit battlefield and collect technical as well as tactical data. The Indian army does have institutions like the faculty of studies at College of Combat or the section 6 of the Military Operations Directorate. But in case of the former, it essentially uses data/information provided by the formations. Unfortunately, the MO-6 has been defunct for a long time. The army training command could be one such organisation, but thanks to the turf wars between formation commanders and training command, this has not come into existence.
The skepticism about data and any kind of statistical analysis is founded on the premise that each tactical/strategic situation is unique. This is indeed quite true as the variables like terrain, weather and human factors like morale (own and enemy’s), leadership and motivation, indeed make each tactical situation ‘unique’. But even accepting these limitations, certain unchangeable like weapons and their effect as well as basic tactical features (fire and move) remain unchanged.
It has been widely accepted that lack of indigenization of our weapon systems is one of our major weakness. This is mainly due to the fact that weapon designs and QRs (qualitative requirements) floated by the services are often based on individual commander’s perception or worse on the current ‘fashion’ in the advanced countries.
Two concrete examples come to mind. Nearly 50 years ago, when artillery support was often in short supply, a very senior General used to often preach that the infantry should plan its attack ‘only’ on fire support of its own mortars. Artillery fire was to be treated as bonus, currently that honour belongs to close air support. All this despite a clear statistical evidence that most battles were won due to gun fire. Even the adoption of 5.56 caliber rifle (INSAS) had its genesis in the Vietnam war orthodoxy (on the American side) that it was better to have personal weapon that could ‘wound’ rather than ‘kill’ an enemy soldier. Is it any wonder that the INSAS rifle has not takers today!
With the proliferation of recording devices, I am aware that virtually every tactical action in our on going counterinsurgency operations, is video graphed. This has rich real time data available for tactical analysis. This should be supplemented with formation level teams to reach the spot of action and interview the participants about the just concluded operation.
The American research agency, DARPA (Defence Advanced Research Project Agency) has an on going practice of inviting battlefield leaders of the ranks of captains and majors to come and interact with scientists on weapon performance/requirements. This is a far better method than the QRs being cooked up in a South Block office.
The DRDO should be consulted on type of data that is needed for developing future weapons. For the tactical data, Infantry College and College of Combat should get together to formulate questionnaire’s for the tactical analysis teams.
It is time we shed the ‘personality’ based approach by data based approach for innovation, both tactical and technical.