Where do Pakistan stand in AI race?
The emergence of Artificial Intelligence (AI) has left a positive influence on the world. Let it be the complex task of image analysis, text translation, speech recognition, and so forth. The aspect which makes AI relatable and interesting is its proof of work which is mostly not observed in the field of Cloud/Edge computing and Blockchain. A wide range of industries have adopted and transformed themselves with the use of AI, let it be Amazon, Netflix, Google, Microsoft, and many more. I can write about the feats which AI has achieved in recent years from beating humans in games to autonomous self-driving cars but that’s not the point of this article. This column aims to let people know where we as a nation stand in the race of adopting AI technologies and on which of its aspects our government needs to focus. Let’s explore a few areas where AI can help in improving the country’s economy.
1. Tourism
Have you heard about “AirBnB” [1] or similar applications? If not, then it’s time to make our own. Our northern areas are the hub of tourist spots both for the nationals and foreigners but do the locals get enough benefit? Can’t we just make a list of the local peoples sharing their rooms, food, and utensils for a couple of days in exchange for a reasonable rent? Most of the foreigners when visiting a country like to experience the traditions, culture, and hospitality along with the spot which they want to visit. This will not only propagate a positive image of our country which has been represented by some Instagram travel influencers visiting Pakistan recently but also benefit the locals and in turn the economic cycle (in the form of taxes).
Now how can AI help? Considering we have a list of locals wanting to share their place along with some attributes like location, nearby tourist spot, food, availability of internet, room size, and so forth, AI can recommend these residential places to the national and international tourists planning to visit in advance. Now the complexity of the recommendation system varies with the preferences, like availability of windows showing scenic views, sharing of a kitchen, availability of Internet connection, and so forth. There are two main benefits of using AI for tourism in Pakistan. The first is of course the financial benefit which can be observed through tax returns and the second is the sustainability of the ecosystem which is caused due to the continuous construction of hotels and buildings in the beautiful northern areas. We have seen what the constructors have done to the city of Murree in the last 20 years. Now let’s move to the next area which is:
2. Farming
It is known globally that Pakistan is an agricultural country and how many measures have been taken to improve the farming process through the use of technology? Well, this question is enough to let the government think about what has been done so far. I assume that everyone in our country takes the agricultural aspect for granted. With the rising population of our country, it would be a challenge for the farmers to feed the population with the available land. Furthermore, this challenge is stacked by weather conditions, and threats like pests, and weeds. These challenges heavily affect the quantity and quality of food. It has been proven by the analysts at the Swiss Federal Institute of Technology (EPFL) and Penn State that the smart machines can help farmers to overcome some of the challenges with consistency, efficiency, and precision [2]. For instance, EPFL gathered 53k images of unhealthy and healthy plants to construct an autonomous differentiating system between infectious and non-infectious plants. The system developed was quite accurate and differentiated the plants with a precision rate of almost 99%. Similarly, Blue River Technology [3], developed a system with help of AI integrated cameras and robots to drive and intelligently spray through fields such that it only targets the weeds while keeping the crops intact. With a similar stance, a Berlin-based startup developed a system (Plantix) [4], that can find nutrient deficiencies and potential defects in the soil through image analysis.
Canadian based SkySquirrel technologies [5] use Drones to perform crop analysis and a Colorado-based company [6] uses AI to evaluate farms, analyze crop sustainability in relation to the weather prediction. Being an agricultural country, shouldn’t we be focusing on and developing such systems while integrating AI? That is the question which the government needs to ask for now as agriculture is one of the sectors which can help retain economic development.
3. Healthcare
With the spread of COVID-19 infection, it just not wreaked the economic industry but also tested the healthcare facilities and affected the population with psychological diseases to a great extent. One of the setbacks in terms of handling COVID-19 infection is the scale at which tests were conducted. It has been proven widely that RT-PCR testing has a sensitivity of 91% [7], yet we have solely relied on the PCR testing so far. The world has moved towards the potential triage methods for increasing the scale of tests such as CT-scans, chest X-rays, Antigen tests, and so forth. Considering the lack of experts, AI can be used extensively with medical images to separate the potential COVID-19 patients from normal or bacterial pneumonia affected ones through analyzing CT-Scans and chest X-rays. However, our government had not explored any options other than the RT-PCR. Moreover, with social distancing and isolation, many patients have developed symptoms of stress, depression, anxiety, and so forth. Smartphones and smartwatches have been the primary accessories people wear and use to perform communication. Exploring the capabilities of these wearable devices beyond their communication capabilities not only can help researchers and doctors to identify stress and depression but also help them to analyze their behavior patterns which elicit such stressful conditions [8–10]. Even further, these wearables can also help in finding physical activity anomalies for sick and elders living independently [8].
These wearables consisting of galvanic skin response and heart rate variability can be further used for detecting emotions and suicidal intentions as suggested in many of the research studies [9], [11]. Another area that has not been explored with reference to COVID-19 is the identification of viral strains. A recent study from King’s College London [12] suggested that there are six distinct types of strains that can be distinguished and assigned to a cluster of symptoms. The analysis of these strains can also help predict how unwell a patient becomes. Unfortunately, neither our government has focused on analyzing viral strains nor adapted any AI system to cluster a particular patient. So far, I have only discussed the AI techniques in relation to the COVID pandemic but moving past from this AI can be used in a variety of medical applications such as Diabetic retinopathy, health monitoring through wearable as well as infrastructure-based sensors, visual monitoring of patients, and many more.
I can go on with the applications of AI in stock markets, surveillance industry, Education, personal fitness, entertainment, trade analysis, corruption control, anomaly detection in government processes, and so forth, but it will transform this article in a short book of a sort. I think will limit my applications for now to three domains.
References:
[1] Airbnb.com
[3] http://www.bluerivertechnology.com/
[5] https://www.f6s.com/skysquirreltechnologiesinc
[7] Xie, X., Zhong, Z., Zhao, W., Zheng, C., Wang, F. and Liu, J., 2020. Chest CT for typical 2019-nCoV pneumonia: relationship to negative RT-PCR testing. Radiology, p.200343.
[8] Khowaja, S.A., Prabono, A.G., Setiawan, F., Yahya, B.N. and Lee, S.L., 2018. Contextual activity based Healthcare Internet of Things, Services, and People (HIoTSP): An architectural framework for healthcare monitoring using wearable sensors. Computer Networks, 145, pp.190–206.
[9] Prabono, A.G., Setiawan, F., Khowaja, S.A., Yahya, B.N. and Lee, S.L., 2017. ETL Workflow Design for Analyzing Relationship of Empathy and Social Awareness using Socio-Emotion Data. 한국정보과학회 학술발표논문집, pp.257–259.
[10] Khowajao, S.A., Prabono, A.G., Setiawan, F., Yahya, B.N. and Lee, S.L., 2018. Towards Anomaly Detection System using Wearable Sensors. Korea DataBase Conference, pp. 1–4.
[11] Setiawan, F., Khowaja, S.A., Prabono, A.G., Yahya, B.N. and Lee, S.L., 2018, July. A framework for real time emotion recognition based on human ans using pervasive device. In 2018 IEEE 42nd Annual Computer Software and Applications Conference (COMPSAC) (Vol. 1, pp. 805–806). IEEE.
[12] https://www.weforum.org/agenda/2020/07/covid-19-study-6-strains-severity-kings-college-london/
Author’s Bio
Dr. Sunder Ali Khowaja received the Ph.D. degree in Industrial and Information Systems Engineering from Hankuk University of Foreign Studies, South Korea; M.E. degree in Communication Systems and Networks; and B.E. degree in Telecommunication from Mehran University of Engineering and Technology, Pakistan. He is currently an Assistant Professor at Department of Telecommunication Engineering, Faculty of Engineering and Technology, University of Sindh, Jamshoro, Pakistan. He had the experience of working with multi-national companies as Network and RF Engineer from 2008 to 2011. His research interests include Data Analytics and Machine Learning for Ambient Intelligence, Image Analysis and Computer Vision applications.
Github page: https://github.com/sander-ali
Publications: https://scholar.google.com/citations?user=JG9e8TsAAAAJ&hl=en&oi=ao
