By Dipankar Roy
The human brain is one of the greatest wonders of creation and its potential is still being explored by medical science. As to how the brain senses the environment and makes the most accurate decisions instantly is still a mystery. The more trained a human brain is the better it can function. When a child realizes that the vapor coming out of a hot soup burns, its brain stores the message for future reference to prevent the child from getting burnt one more time. To the brain, the input here is a ‘vapor image’ and output the ‘pain’. Artificial Intelligence and Machine Learning concepts are likewise like the brain. The higher the training, the closer it approaches from probability towards science.
Until recently, Artificial Intelligence wasn’t a well-accepted concept but it was looked upon as an add-on capability to aid in business suggestions. Artificial Intelligence has come a long way; today it is not just for suggesting business ideas but has become a breakthrough technology enabling us to take business decisions cautiously learning about their pros and cons.
According to a report by Boston Consulting Group, Artificial Intelligence will boom to a 100 billion market by the year 2025 with heavy corporate investments likely this year. Previous year has seen around 5 billion being pumped in as venture investment in machine learning.
Combination of “pre-defined rules” by business experts and “generated rules” by amazing algorithms derived from historical data is the heart of Artificial Intelligence. Concepts like Decision Tree, Clustering, Association Analysis, Pattern and Image Recognition, Neural Networks, Text Mining have the potential to bring out wonders. Software companies need to come together and invest in artificial intelligence to touch the lives of billions positively. Artificial Intelligence and Machine Learning are the game changers and have further potential to create wonders, a few listed below.
Organization Efficiency:
Talent is abundant, but key to organizational efficiency highly depends in acquiring the right talent. Starting from resume matching techniques using text mining to simulating interview panels with a combination of artificial intelligence, decision trees, text mining and voice to text integration, an organization can achieve a fast, efficient and unbiased hiring process.
On the execution efficiency front, various parameters like topic category, availability, ad-hoc influence factors, customer incident data, project schedules, company holiday calendar, etc. in a project can be analyzed to recognize the improvement areas.
Sales Forecasting and Customer Retention:
Mining of customer actions, sales transactions, regionals sales distribution, sales executive involvement, social sentiment, customer incident data together with time series can be used to forecast sales, recognize the positive influencers and identify customers who are at high risk of leaving. Artificial Intelligence plays a major role for businesses involving sales and customer relations.
Such actions could be involvement of right sales executive, general promotions, regional promotions, customer specific offers.
Supply Chain and Transportation Management:
The production forecasts based on order received are greatly influenced by operational obstacles in the different stages of supply chain management. These obstacles act as deterrent to the profitability of the industry. The supply chain and management involves the management of flow of goods and services, raw material storage, inventory, and delivery of goods.
Whole process is critically dependent on quantity of order received, lead time, procurement time, quality standards and inventory level of warehouses, manufacturing assets.
Factors like breakdown, re-planning frequency, BOM complexity, Procurement delays, are factors affecting timely delivery. In combination with time series data like regional weather conditions impacting route planning, seasonal product demand, government regulations – techniques such as clustering, decision tree, regression can be used to accurately plan supply chain, recommend transportation channels and packaging and predict costs.
Image processing techniques could be used to predict the manufacturing asset situation and help preempt losses. For example, deviation in a nut-bolt image from standards can indicate that the chamfer-taper needs maintenance. The system can be artificially intelligent to stop further processing through the defective equipment and raise a maintenance order.
Apart from these, artificial intelligence and machine learning techniques have been high value add to healthcare, retail execution, career planning, personalized education, fraud detection and financing.
As software organizations come close to research in various unknown frontiers, we are slowly reducing the gap of the artificially intelligent machines and the human brain. Although, the way our brain can perform in entirety remains a mystery, we would at least reach a state where we simulate some of its capabilities very accurately. This increased efficiency through artificial intelligence gives us bandwidth to think of more complex problems and find the right solutions to positively touch the billion lives in various spheres.
The author is development manager at SAP Labs India. The view expressed by the author is his own and does not reflect the views of the company