AI for Organisations in Malawi

How to Spot Opportunities to Apply AI to Problems in your own Organisation

Amelia Taylor�Malawi University of Business and Applied Sciences�KuyeseraAI Lab

Access this PPT at: https://shorturl.at/cdvz1

Icebreaker

• Can you think of ways in which Digital Transformation and AI is currently impacting your life/ work / organisation?

​

​

• Can you think of ways in which Digital Transformation and AI will impact your life / work / organisation in the next 5 years?

A. Taylor, MUBAS

Access this PPT at: https://shorturl.at/cdvz1

Icebreaker

• Connected cars: integrated AI technology to facilitate communication between “cars and cars" and "cars and infrastructure (road equipment)".
• Mobile Technology: AI features for authentication
• 5G Surveillance: already running on PCs and mobile phone send a lot of data to renters such as Google, Huawei, Microsoft; 5G will result in integrating in real time a multitude of data points.

A. Taylor, MUBAS

Access this PPT at: https://shorturl.at/cdvz1

Digital Transformation

Systems of intelligence

= Acceleration of business activities, processes, competencies, and models to fully leverage the changes and opportunities of digital technologies and their impact in a strategic and prioritised way.

​

A fundamental change in how companies generate value for their owners and other stakeholders. 

A. Taylor, MUBAS

–Pierre Nanterme, CEO of Accenture

''Digital is the main reason just over half of the companies on the Fortune 500 and disappeared since the year 2000.''

A. Taylor, MUBAS

Access this PPT at: https://shorturl.at/cdvz1

5 Factors for Successful Digital Transformation

• Digital Savy Leaders
• Building Workforce Capabilities
• Empowering Employees to work in new ways (together)
• Give day to day tools an upgrade
• Communicate frequently via digital and traditional methods

A. Taylor, MUBAS

Technology Trends

According to recent Siemens and McKinsey surveys the following current technology trends will drive more digital transformation:

1. Computing and visual technology
2. Connectivity
3. AI
4. Product lifecycle: ideation, realisation and utilisation

A. Taylor, MUBAS

Computing and visual technology

• Advances in hardware: computing power continues to grow.
• Flexibility:
• Hybrid Cloud: hybrid cloud provides economies of scale while simultaneously delivering security.
• Edge Computing: McKinsey study claims that 127 new IoT devices connect to the Internet every second. Gartner states that by 2025, 75% of enterprise-generated data will be created and stored at the edge.
• Hyper-converged Infrastructure - IT framework that integrates storage systems, servers, and networking into one single platform (private cloud).
• The use of AI in Data Centres: energy saving, fault detection (Malawi Data Center in Blantyre)
• The use of containers: the International Data Corporation (IDC) more companies will use containerised applications in production.

A. Taylor, MUBAS

Access this PPT at: https://shorturl.at/cdvz1

Connectivity

5G Communications/ IoT / Electronics and Semiconductors

Connectivity technologies enable digital enterprises to be more competitive by:

• Escaping the constraints of geography
• Collecting more information
• Responding to customers more quickly
• Improving productivity and efficiency
2. (Advances in connectivity is sometimes called The invisible revolution)
• working from home / learning from home
• optical cables - connecting countries / cities (Malawi National Fiber Optical Backbone project - how to leverage this?)

A. Taylor, MUBAS

AI

• Symbiotic relationship between Digital Transformation and AI
• 4 approaches to AI : Acting humanly, thinking humanly, thinking rationally, acting rationally

A. Taylor, MUBAS

AI Technology: What is it?

• In general, AI is a system’s ability to:
• Correctly interpret external data.
• Learn from such data.
• Use those learnings to achieve specific goals and tasks through flexible adaptation.
• Constantly adapt to internal and external conditions and requirements.

• There are four parts of the AI System, each with individual roles, that should be considered:
• Artificial Intelligence - Ability to sense, reason, engage and learn
• Machine Learning - Ability to learn (unsupervised, supervised, and reinforced)
• Methods - Ability to reason
• Technologies - Physical enablement

A. Taylor, MUBAS

AI in Production

Image Recognition / Speech recognition / Chatbots / Game playing

Health Care:

- Quality Assurance: identifying antibiotic resistance

- Image recognition for precision diagnosis of medical conditions

- Robotic surgery

- Digital health: global demographics and public health, use of IoT

Manufacturing:

- Quality Assurance: identify defects based on data

- Digital twins

A. Taylor, MUBAS

Product Life-Cycle

• Digital twins: a virtual imitation of an actual process, product or service occurring in the real world. In other words, it takes real world data about a physical object or system as inputs and then produces as outputs, predications or simulations of how the physical object or system will be affected by those inputs.
• Visual technology - VR, AR

A. Taylor, MUBAS

Digital Transformation & Maturity

Deloitte Study:

Higher-maturity organisations’ digital transformation efforts tend to be twice as broad as those of lower-maturity organisations.

Malawi: Data and Technology - 1

• Stakeholders: Government / Education / Research / Business / Manufacturing / Finance / Telecoms
• Government Digitisation projects: National ID cards, Malawi Traffic Information System, Integrated Fiscal Management and Information System, Automated System for Customs, Unified Beneficiary Registry, Personal Property Security Registry System, Judiciary Case Management Information System, Virtual Landing Point, Passport System.
• Telecoms: ICT Infrastructure (National Fiber Backbone project/ ESCOM), New systems or pricing schemes for internet access, mobile money, village banking

17

A. Taylor, MUBAS

Malawi: Data and Technology - 2

• Finance: significant digitisation and large capital, internet banking / mobile money / integrated payments; how about advanced technologies such as ML, Data Mining, IoT?
• Academia: new universities / curriculum with topics such as IoT, drones, AI, ML / new programs at MSc / academies - drone, Huawei, support networks - ICTAM, Innovation Hubs, IndabaX, PanAfrican Network, Mzuni Data Science Bootcamp
• Business:
• Manufacturing:

18

A. Taylor, MUBAS

Malawi: Data and Technology - 3

• Enabling Factors: Internet Infrastructure / Access and Use / Education / Research
• Internet Access and Use: Trend is mostly UP (young, male, urban, ownership pf mobile phone, mobile broadband, usage & access gaps, high costs) => workforce more connected and tech-savvy => but international bandwidth usage gap is widening between LDCs and World.
• Digital Development: ICT ecosystem (vision: how your efforts contribute to the larger vision and strategy? Low visibility of the players in the ecosystem?)

19

AI: How to get from Point A to Point AI?

4Ds (Source intel.ai)

• Discover (identify and prioritise, consider and organise),
• Data: what is a data strategy? What does it contain? From Data to Insights / ingest/ store / prepare
• Develop (set up/model/test and document): AI algorithms involve a lot more trial and error
• Deploy: digital transformation, cloud options, automation

A. Taylor, MUBAS

Case Study 1: COVID-19 Infodemy

• Plan: Build an ‘intelligent’ catalog of articles (news and journal publications) on Covid-19 in Malawi and Kenya
• Intelligent because: (1) it can automatically detect articles of interest add them to the catalog and (2) annotate them with keywords and (3) extract summaries of their content.
• Usage: analyse the flow of information during the pandemic, contextualise public health measures in terms of public reaction, understanding this also findings from medical studies in terms of diagnosis, effectiveness of these measures.

A. Taylor, MUBAS

Case Study 1: COVID-19 Infodemy

• How did we go about that?
• STAGE 1:
• Journal Articles: used APIs and manual search
• News: manually identified sources of news and articles by searching
• Process of searching was formalised/controlled: e.g. we documented the keywords selection, creates a gold standard
• Developed an online interface: https://c19na.vercel.app/news
• STAGE 2: develop an automated news harvester by leveraging the cloud infrastructure (Macphail Magwira)

A. Taylor, MUBAS

Case Study 1: COVID-19 Infodemy

• How did we go about that?

​

• STAGE 2: develop an automated news harvester by leveraging the cloud infrastructure
• Demo: Macphail Magwira

A. Taylor, MUBAS