New Delhi: Effectiveness of Artificial Intelligence (AI), machine learning, robotics and cognitive automation in direct proportion bestow rise in the quality and quantity of training data that the systems are exposed to, the conditions are ripe for India to emerge as a leader in AI, said an ASSOCHAM-PwC paper.
India is already on the path of a digital revolution and the next step is utilising the big data generated to take intelligent decisions. This requires close collaboration between academia, the private sector and public sector in order to understand problems holistically and solve them, , the ASSOCHAM-PwC joint study on ‘Artificial Intelligence and Robotics – 2017’ noted.
AI augmented manufacturing operations can employ more reliable demand forecasting, a flexible and responsive supply chain, quicker changes in operations, and more accurate scheduling and inventory optimisation. Other benefits involve creation of smarter, quicker and environmentally sound processes. This can lead to increased productivity and quality, lower costs and a more robust health and safety framework.
The application of AI in the field of defence and security includes protection of infrastructure such as airports, power plants and economic sectors that are vulnerable to attacks, detecting anomalous behaviour in individuals, and using distributed sensors and pattern recognition to predict infrastructure disruptions through natural/man-made causes, adds the study.
The ‘security games framework’ is based on computational game theory, combined with elements of human behaviour modelling. Given the limited security resources and different high value targets, game-based decisions provide a randomised collection or patrolling schedule based on weights of targets and intelligent reactions of adversaries to security postures.
AI shows remarkable potential in aiding control and remedial actions in the aftermath of environmental and man-made disasters. It can assist in optimising mobile networks and smart bandwidth allocation to ensure network service continuity in the midst of catastrophic events that are usually followed by a spike in communication and jammed networks.
Unmanned drones and satellite feeds combined with image processing and recognition can be used in infrastructure damage assessments and predictions based on structural stability and traffic congestion avoidance through adaptive routing while equipping and deploying disaster management teams. Opportunities for AI intervention also reside in processing social media feeds to gauge location-specific urgency and send targeted alerts to minimise loss of life and property, pointed the study.
A key area of AI intervention in logistical operations involves adaptive scheduling of deliveries and routing of vehicles. Advanced logistics and supply chains are being created using expert decision systems. Products can be transported more efficiently through vision-based driver assist and automated/robotic systems. This has made transportation less susceptible to disruptions caused by weather, traffic and unnatural events.
Some of the major areas of application of AI in the banking and financial services sector include early detection of financial risk and systemic failures, and automation to reduce malicious intent in financial systems, such as market manipulation, fraud, anomalous trading and reduction in market volatility and trading costs.
AI can improve the efficiency of operations in the travel and transportation sector by bringing improved safety through structural health monitoring and infrastructure asset management that can pay dividends in terms of reduced cost of repairs and reconstruction and real-time route information, thereby reducing energy usage and emissions.
Agriculture is another sector that can greatly benefit from intelligent solutions by using smarter production, processing, storage, distribution and consumption mechanisms. AI solutions can also help provide site-specific and timely data about crops to enable application of appropriate inputs such as fertilisers and chemicals, highlighted joint study.
Consumer goods and services was one of the initial areas of AI adoption in India and currently accounts for a significant share of private sector application. To enable consumers to find better products at low prices, machine learning algorithms are being deployed for better matching of supply with consumer demand.
Efficient usage of bandwidth and storage, improved filtering, web searches and language translation are some of the benefits of employing AI systems in the communication and social media sector. AI can enhance scientific research and experimentation by assisting scientists and engineers in reading publications and patents, generating hypotheses and testing them through the usage of robotic systems.
Large parts of the country experience a dearth of academicians and teachers when it comes to making education effective for students across a gradient of social and cognitive abilities. AI solutions can meaningfully intervene by means of adaptive tutoring based on the receptiveness of students and accurate gauging of development of students complemented by in-person classroom learning.
Evidence-based treatment and medication require a level of precision that helps patients develop confidence and trust in their doctors—an area where mere manual experience and judgment may be supplemented with AI.
With the vast volume of information-processing capabilities required for fields such as bioinformatics, using AI-based algorithms and solutions is inevitable. AI application in healthcare, medicine and biotechnology includes supporting systems to identify genetic risks from large-scale genomic studies, predicting safety and efficacy of newly launched drugs, providing decision support for medical assessments and prescriptions and tailoring drug administration to the individual patient.
Some of the areas where AI can improve legal processes include improved discovery and analysis based on law case history and formulation of legal arguments based on identification of relevant evidence.
Cognitive technologies are being deployed by firms to largely automate the task of going through stacks of documents to identify key terms, which has until now been a time-consuming manual process. NLP technology reads and understands key points in the documents. Machine-learning technology makes it possible to train the system on a set of sample documents so that it learns how to identify and extract information in an automated manner.