Use artificial intelligence, machine learning for efficient relief management

  • Hannah
  • September 9, 2020
  • Comments Off on Use artificial intelligence, machine learning for efficient relief management

Bangladesh located to the north-east of South Asia is one of the disaster-prone areas of the region as well the world due to its geographical characteristics, complex topographical features, excessive and mighty river system, monsoon climate and the coastal morphology.

This in turn makes the country’s population susceptible to many natural hazards, such as floods, tornados, salinity intrusion, droughts and earthquakes.

Between 1972 and 2019, Bangladesh witnessed over 300 natural disasters. 

The two most common natural disasters that hit the country are floods and cyclones, accounting for 47.47 per cent and 28.96 per cent of the total disasters respectively.

However, a humanitarian crisis unlike any other arose with the advent of Covid-19. The novel coronavirus has already dealt a significant blow to the society in general and the global economy, to which Bangladesh belongs.

According to a study, among the 100.22 million people that are at risk both economically and health-wise, 53.64 million of them are extremely poor.

On average, natural disasters have a 1.8 per cent bump in Bangladesh’s GDP every year and on top of that, this year the losses brought about by the ongoing pandemic are significantly higher than before.

During any crisis, the top priority for any government is to run their relief operation efficiently so that aid can be distributed to those in need on time.

Amid any such situation, many of us think of providing immediate financial support to all affected people. However, one should remember that Bangladesh’s economy has reached a level where the government is fully capable to handle the issue with their own funds.

Bangladesh’s GDP witnessed an annual growth of about 7.76 per cent in the last three years. By the end of 2019, the total GDP stood at $305 billion, which is one of the highest in the world.

At the end of July this year, the country’s forex reserve reached a record $ 37.10 billion — 52nd in global ranking and second among the Saarc nations. As per the latest ADP forecast, Bangladesh will post the highest annual GDP growth in Asia in the ongoing fiscal.

Over the years, Bangladesh has earned international recognition for its efficient relief management operations, but still there is plenty of room for improvement.

In Bangladesh, one of the biggest challenges is to identify the right target group or affected people. Even now, we need to manually identify the groups of people in need of aid and because of this the allocated fund usually fails to reach the intended beneficiaries.

Recently, the government has decided to provide 50 lakh poor families impacted by the Covid-19 crisis cash assistance of Tk 2,500 each. But since those in need were not properly identified, a significant portion of fund remained unutilised.

To track the right group of people, a cut in lead time and cost, increased transparency, use of artificial intelligence (AI) and machine learning (ML) technology and effective supply chain management could play an active role in that regard.

Active role of AI and ML for digitalisation of database

This is the right time to work on a technology-based comprehensive national database for successful disaster management. By using AI and ML support, we can overcome this limitation.

We have many sources of data, such as around 130 social safety net programmes that are run by the government. The number of active mobile users stands at 162.92 million while 92.57 million people are registered mobile financial services (MFS) account holders. The monthly number of MFS transactions is about 310 million.

By analysing user transactions, the concerned authorities can identify a person’s financial position along with his/her purchasing power and geographic location.

However, this will only be possible if we use AI, ML and other data sources as obtained from the Bangladesh Election Commission and Bangladesh National Bureau of Statistics.

The social welfare ministry and the disaster management and relief ministry need to form a regulatory body by involving all relevant ministries, Bangladesh Telecommunication Regulatory Commission, Bangladesh Bank, Bangladesh Army, representatives of mobile financial service providers, mobile operators, developing partners and leading non-governmental organisations to develop a digital database.

The body would also extend assistance during any sort of disasters or pandemic and should have a proactive approach rather than a reactive one.

There are many other fields where AI and ML are beginning to have clear impacts on disaster preparedness and relief distribution.

Image and information support can be explored using the Bangabandhu Satellite-1 to build a ML model that could estimate the damage to all affected areas and infrastructures in order to reduce the amount of human labour and time required to plan an appropriate response.

There are three main types of relief: food, essentials and cash support. During and post-disaster relief operations and effective supply chain management mainly consist of efficient handling of multiple supply chain activities, such as on time cost-effective procurement, faster stock fulfilment, safe warehousing facilities and shorter lead times.

On time cost-effective supply

According to an extensive study, there are mainly two types of supply chains operated by the government to assist distressed people. The first is to assist poor people throughout the year on humanitarian grounds while the second is operated during any disaster.

The first one is well planned and can be considered as a part of the government’s routine humanitarian activities but the second is unplanned and sudden.

Supply chain activities for disaster relief management has plenty of room for improvement.

To ensure cost-effective and on time delivery, a disaster relief procurement committee could be formed with representatives from all government and non-government bodies. The new body should hold regular quarterly meetings, focusing on effective supply chain initiatives.

Next, based on standardised food and non-food items as per the last couple of years’ requirement trend analysis, comprehensive study on procurement activities should be recommended before finalising a schedule of rate.

This will not only reduce the procurement lead time but will also reduce the procurement cost.

Stock fulfilment and warehousing facility

Based on last year’s disaster trend analysis, it is recommended for selected district level warehousing/storage facilities to store all food and non-food items in a centralised warehouse instead of in different locations as that will actively contribute to the reduction of holding costs and enhance operational efficiency.

Logistics excellency through reduction of delivery lead time

During and after disasters, one of the key reasons behind delays in relief management are the poor and non-structured logistics operations.

Proactive initiatives and effective logistics operations will not only accelerate relief distribution, but also ensure better transparency along with cost control.

First of all, a pre-defined logistics committee should be formed with representatives of all relevant government bodies, and, if required, relevant non-governmental organisations.

They should assess the requirement and take on-time proactive initiatives by analysing the trend of last couple of relief operations.

Secondly, they should have good coordination with other government agencies such as postal services, BRTC, BADC, Bangladesh Railway along with exploring the use of their transport facilities to not only to reduce logistics cost, but also to accelerate relief operations.

Distribution of cash support

Among developing countries, financial inclusion in Bangladesh is one of the best. bKash, a local company that plays a major role in this respect, is now the biggest MFS organisation in the world and the country now has around 88.8 million MFS accountholders, according to Bangladesh Bank data.

Cash disbursement support through MFS services will not only reduce lead time, but also ensure transparency and safe holding of cash by the end-users. This will also help to develop a record that could be used for future analysis, with the best example being the government’s latest cash support scheme to support poor families amid the ongoing pandemic.

The present government aims to present a beautiful lifestyle for the people by ensuring their basic needs. Furthermore, their vision is to establish Bangladesh as a poverty- and hunger-free country by 2021 and a developed one by 2041.

The author is head of supply chain and procurement at bKash.