Artificial Intelligence


Artificial intelligence is not new, yet there have been quick advances in the field as of late. This has to a limited extent been empowered by improvements in processing power and the colossal volumes of advanced information that are presently produced. A wide scope of utilizations of AI are currently being investigated with significant open and private speculation and premium. The UK Government reported its aspiration to make the UK a world head in AI and information advancements in its 2017 Industrial Strategy. In April 2018, a £1bn AI part bargain between UK Government and industry was reported, including £300 million towards AI research. AI is commended as having the capacity to help address significant wellbeing challenges, for example, meeting the consideration needs of a maturing populace. Significant innovation organizations – including Google, Microsoft, and IBM – are putting resources into the improvement of AI for human services and research. The quantity of AI new businesses has likewise been consistently increasing. There are a few UK based organizations, some of which have been set up as a team with UK colleges and clinics. Organizations have been framed between NHS suppliers and AI engineers, for example, IBM, DeepMind, Babylon Health, and Ultromics.

Healthcare Organization – Artificial intelligence can possibly be utilized in arranging and asset assignment in wellbeing and social consideration administrations. For instance, the IBM Watson Care Manager framework is being guided by Harrow Council with the point of improving cost productivity. It matches people with a consideration supplier that addresses their issues, inside their distributed consideration spending plan. It additionally structures singular consideration plans and claims to offer bits of knowledge for increasingly successful utilization of care the executive’s resources. AI is likewise being utilized with the point of improving patient experience. Birch Hey Children’s Hospital in Liverpool is working with IBM Watson to make a ‘psychological medical clinic’, which will incorporate an application to encourage collaborations with patients. The application means to distinguish persistent tensions before a visit, give data on request, and furnish clinicians with data to assist them with delivering suitable medications.

Medical Research – Artificial intelligence can be utilized to dissect and distinguish designs in enormous and complex datasets quicker and more decisively than has recently been possible. It can likewise be utilized to look the logical writing for pertinent investigations, and to consolidate various types of information; for instance, to help sedate discovery. The Institute of Cancer Research’s jars AR database joins hereditary and clinical information from patients with data from logical research and uses AI to make forecasts about new focuses for malignancy drugs. Researchers have built up an AI ‘robot researcher’ called Eve which is intended to make the procedure of medication disclosure quicker and more economical. (K.Williams, 2015) AI frameworks utilized in human services could likewise be significant for restorative research by coordinating reasonable patients to clinical examinations.

Clinical Care – Artificial intelligence can possibly help the analysis of illness and is presently being trialed for this reason in some UK emergency clinics. Utilizing AI to investigate clinical information, examine distributions, and expert rules could likewise advise choices about treatment


APPLICATIONS – A few applications that utilization AI to offer customized wellbeing appraisals and home consideration exhortation are as of now available. The application Ada Health Companion utilizes AI to work a talk bot, which joins data about side effects from the client with other data to offer conceivable diagnoses. GP at Hand, a comparative application created by Babylon Health, is as of now being trialed by a gathering of NHS medical procedures in London. Information devices or visit bots driven by AI are being utilized to help with the administration of constant ailments. For instance, the Arthritis Virtual Assistant created by IBM for Arthritis Research UK is learning through associations with patients to give customized data and guidance concerning prescriptions, diet, and exercise. (Release, 2017) Government-financed and business activities are investigating manners by which AI could be utilized to control mechanical frameworks and applications to help individuals living at home with conditions, for example, beginning time dementia. Man-made intelligence applications that screen and bolster tolerant adherence to recommended drug and treatment have been trialed with promising outcomes, for instance, in patients with tuberculosis. (L.Shafner, 2017) Other apparatuses, for example, Sentrian, use AI to examine data gathered by sensors worn by patients at home. The point is to identify indications of decay to empower early mediation and avoid medical clinic affirmations.

PUBLIC HEALTH – Artificial intelligence can possibly be utilized to help early location of irresistible malady flare-ups and wellsprings of pandemics, for example, water contamination. (B.Jacobsmeyer, 2012) AI has likewise been utilized to anticipate unfavourable medication responses, which are assessed to cause up to 6.5 percent of emergency clinic affirmations in the UK.

Babylon a UK fire up plans to “put an open and reasonable wellbeing administration in the hands of each individual on earth” by putting man-made brainpower (AI) apparatuses to work. Right now, the organization has activities in the UK and Rwanda and plans to extend to the Middle East, the United States, and China. The organization’s technique is to consolidate the intensity of AI with the medicinal aptitude of people to convey unrivalled access to human services.

How does Babylon’s AI work?

A submitted group of research researchers, architects, specialists and disease transmission experts are cooperating to create and enhance Babylon’s AI capacities. A great part of the collaboration is on the advancement of bleeding edge AI explore; this is being passed through access to enormous volumes of information from the therapeutic network, constant gaining from our very own clients and through input from Babylon’s very own specialists.

The knowledge graph and user graph:

Babylon’s Knowledge Graph is one of the biggest organized medicinal information bases on the planet. It catches human information on present day medication and is encoded for machines. We utilize this as the reason for Babylon’s clever parts to address one another. The Knowledge Graph monitors the significance behind therapeutic phrasing crosswise over various restorative frameworks and various dialects. While the Knowledge Graph gives the general information about medication, tolerant cases are kept in the User Graph. Conjunction of the Babylon Knowledge Graph and the User Graph takes into consideration more revelation. We can coordinate indications with data and results, continually improving the data we give.

The inference engine:

Essentially seeing how clients express their indications and hazard factors isn’t enough to give data on perhaps coordinating conditions. At the core of Babylon’s AI is our surmising motor, an amazing arrangement of AI frameworks, equipped for thinking on a space of >100s of billions of blends of indications, illnesses and hazard factors, every second, to help distinguish conditions which may coordinate the data entered. The surmising motor gives our AI the capacity to give thinking productively, at scale, to carry wellbeing data to millions.

Natural Language Processing (NLP):

Our AI can’t give data to patients on the off chance that it can’t get them, and patients won’t utilize our AI if they can’t get it. To help cross over any barrier, we utilize Natural Language Processing (NLP). NLP enables PCs to translate, comprehend, and afterward utilize each day human language and language designs. It separates both discourse and content into shorter parts and deciphers these increasingly reasonable squares to comprehend what every individual segment means and how it adds to the general importance, connecting the event of restorative terms to our Knowledge Graph. Through NLP our AI can decipher counsels, outline clinical records and visit with clients in a progressively characteristic, human way.

Machine Learning research at Babylon:

All through the Babylon stage we use Machine Learning (ML) for an assortment of undertakings. In the induction motor we consolidate probabilistic models with profound learning methods to accelerate the deduction procedure. In the Knowledge Graph we anticipate new connections between medicinal ideas dependent on perusing restorative writing. In NLP we assemble language understanding models dependent on enormous scale datasets of communications with our clients and information from the web. We use ML to show our NLP framework new dialects.

Babylon would not be achievable without the utilization of cutting-edge ML procedures, so we’ve put fundamentally into building a world class inquire about group in this field. Babylon is additionally quick to contribute back to the AI people group through papers, blog entries, and by publicly releasing a portion of our work to help all.

Services Babylon Offers:

Babylon engineers, doctors, and researchers built up an AI framework that can get information about the manifestations somebody is experiencing, contrast the data with a database of known conditions and sicknesses to discover potential matches, and afterward recognize a game-plan and related hazard factors. Individuals can utilize the “Ask Babylon” highlight to ask about their restorative worries to get an underlying comprehension of what they may be managing, yet this administration isn’t proposed to supplant the mastery of a specialist or be utilized in a health-related crisis.

In quest for its strategic, offers a “converse with a specialist” administration by means of its application, GP at Hand that gives day in and day out access to medicinal services experts through video or sound conferencing. The application can be downloaded from Google Play or the App Store. At the conference, specialists can offer medicinal guidance, answer questions, examine treatment, and can arrange solutions that can be conveyed to a patient’s entryway. All the patient’s clinical records are put away in a safe domain, and their wellbeing history can be gotten to and referenced when it’s required. If a patient needs to return to their arrangement, they can audit the restorative notes and replay an account of the arrangement whenever would not be achievable without the utilization of cutting-edge ML procedures, so we’ve put fundamentally into building a world class inquire about group in this field. Babylon is additionally quick to contribute back to the AI people group through papers, blog entries, and by publicly releasing a portion of our work to help all.

Another feature that is available on the app is Healthcheck. Built with the support of doctors, scientists and disease experts, this AI tool can take answers from questions about family history and a person’s lifestyle and compare it to the medical database to then create a health report and insights to help someone stay healthy.

The beginning up claims that in its own tests, the AI framework was spot on80 percent of the time and that the instrument was never intended to totally supplant the counsel of a genuine specialist, yet to decrease holding up times and to assist specialists with settling on progressively exact choices. The world is confronting an outrageous lack of specialists and medicinal experts, and tech, for example, what Babylon offers is one approach to help improve the social insurance of a great many individuals. As indicated by NHS England, “Every security case [of Babylon] satisfies the guidelines required by NHS and has been finished utilizing a hearty appraisal technique to an elevated expectation.”

While it probably won’t be an ideal framework, Babylon shows that man-made reasoning has sufficiently advanced to work nearby medicinal services experts and can be a useful instrument. Be that as it may, patients still need to stay to be their very own furious social insurance advocates. On the off chance that the guidance got from man-made reasoning doesn’t appear to hit the imprint, it’s a word of wisdom to demand a subsequent supposition—from a human.

AI for the patient and provider

Babylon Health needs everybody with a cell phone to approach moderate medicinal services. They accept an application that offers moment conclusion is the key. As their CEO, Ali Parsa, disclosed to the Telegraph: “[Medical professionals] are the costliest piece of medicinal services. What’s more, the second… is timing… [By] the time [most diseases] present their indications a £10 issue has become a £1,000 arrangement.”

Babylon Health accepts they can drop both of those expenses. Today, Babylon Health offers a free application that makes it basic for clients to follow their wellbeing and counsel their AI-controlled chatbot. For a charge, clients can video-visit with top specialists who can get to that client’s wellbeing records and a lot of exclusive AI-fuelled apparatuses that Babylon Health cases can improve treatment quality. By following the vitals, medicines, and results over an expansive client base, Babylon wellbeing has tapped an unbelievably important dataset. This dataset makes it adaptable to consistently improve their AI’s presentation nearby clients’ wellbeing.

IBM Watson for Oncology has a smaller center: improving the results of disease medications. IBM accepts they can give each restorative expert treating disease a similar knowledge that specialists at top malignancy examine focuses have. IBM has banded together with experts at Memorial Sloan Kettering to prepare their PCs with an abundance of restorative records and research. Propelled in 2016, Watson underpins specialists with tolerant explicit suggestions from bleeding edge medications in a small amount of the time. As indicated by Deborah DiSanzo, the General Manager of IBM Watson Health, Watson for Oncology had just been utilized in the treatment of 16,000 patients by the second from last quarter of 2017. With PCs taking care of the examination, specialists can concentrate on what people exceed expectations at: treating the passionate misery of a patient battling malignant growth.

Data for artificial intelligence is food for thought:

Both IBM Watson and Babylon Health concur: specialists can convey better treatment by gaining from the aftereffects of different patients. Computer based intelligence can gain from chronicled information and figure how a patient’s sickness would react to treatment choices. The two organizations are utilizing AI, a strategy that has gotten synonymous with AI lately. AI is a mechanized method utilized by a PC to encourage itself to settle on choices utilizing preparing information. Preparing information is the fuel of AI, as depicted by Andrew Ng of Stanford University.

Babylon Health and IBM Watson have both structured frameworks that produce this “fuel” from their clients. As they draw in more clients, they will create better bits of knowledge. This system impact is a temperate circle where the item turns out to be better as it includes more clients. The drawback of items with organize impacts is that they are famously hard to kick-start. Simply think that it is so difficult to get the initial barely any individuals for a dating site.

Babylon Health and IBM Watson have each collaborated with set up players to defeat this test and get the fuel they must prepare. Babylon Health is bootstrapping their item with assistance from a UK NHS organization. The UK NHS is looking for approaches to relieve their primary care physician deficiency and will preliminary Babylon’s chatbot for a half year in North Central London, a territory covering 1.2 million residents. IBM Watson is cooperating with Memorial Sloan Kettering to help train Watson on the abundance of clinical data and therapeutic aptitude that the middle is known for.

Regulatory risk: A potential challenge:

With AI-fuelled human services items indicating so a lot of guarantee, one may anticipate that guideline should go rapidly through the FDA. In any case, the FDA is right now battling. As the Wall Street Journal puts it:

“How on earth would you say you will manage programming that learns?”

Current guidelines need principles to evaluate the wellbeing and adequacy of AI frameworks, which the FDA has endeavoured to address by giving direction to surveying AI frameworks. The principal direction orders AI frameworks as “general wellbeing items”, which are inexactly managed as they present okay to clients. The subsequent direction legitimizes the utilization of certifiable proof to evaluate the presentation of AI frameworks. In conclusion, the direction explains the principles for the versatile structure in clinical preliminaries, which would be generally utilized in surveying the working qualities of AI frameworks.

Notwithstanding these difficulties, things are looking bullish for AI-fuelled medicinal services. Babylon Health and IBM are just two of numerous new activities that are expanding the range of medicinal services by strengthening the parts that don’t scale: specialists. While every one of these organizations has their very own perspective on the future, they all concur that AI will let our constrained restorative experts carry the best medicines to the best number of individuals. Particularly when the best treatment is acting before we become ill.


Artificial intelligence relies upon advanced information, so irregularities in the accessibility and nature of information confine the capability of AI. Likewise, huge registering power is required for the examination of huge and complex informational indexes. While many are energetic about the potential employments of AI in the NHS, others point to the down to earth difficulties, for example, the way that medicinal records are not reliably digitized over the NHS, and the absence of interoperability and institutionalization in NHS IT frameworks, computerized record keeping, and information labelling. There are inquiries concerning the degree to which patients and specialists are OK with advanced sharing of individual wellbeing data. Humans have properties that AI frameworks probably won’t have the option to genuinely have, for example, compassion. Clinical practice regularly includes complex decisions and capacities that AI as of now can’t imitate, for example, logical information and the capacity to peruse social cues. There is additionally banter about whether some human information is implicit and can’t be taught. Claims that AI will have the option to show self-governance have been addressed on grounds this is a property basic to being human and cannot be held by a machine.

Overall, artificial intelligence advances are being utilized or trialed for a scope of purposes in the field of social insurance and research, including identification of illness, the executives of constant conditions, conveyance of wellbeing administrations, and medication disclosure. Simulated intelligence advances can possibly help address significant wellbeing challenges yet may be restricted by the nature of accessible wellbeing information, and by the powerlessness of AI to have some human qualities, for example, empathy. The utilization of AI raises a few moral and social issues, a significant number of which cover with issues raised utilizing information and human services advances more extensively. A key test for future administration of AI advancements will guarantee that AI is created and utilized in a manner that is straightforward and good with general society intrigue, while animating and driving development in the part.

This article is co-authored by Prof Raul Villamarin Rodriguez, Aakriti Jain, Mohit Mohan Saxena, Epari Shravan and Vaibhav Yadav, Universal Business School.

Machine Learning

Supply Chain 4.0: AI and Robotization

Automatic processes, machine learning, and robotization force a constant updating of the knowledge for those professionals responsible for logistics in MNCs and SMEs.

On one hand, ERP systems have become the nerve center of companies and within these the logistics sector is the true heart and engine of all activity. Specialized logistics services companies proliferate and create models that are responsible for the dynamization of international trade, both B2B and B2C.

These systems transcend the mere management we have known so far, processing thousands of data generated from all departments of the company or collected by automata. This data is systematically analyzed and managed by algorithms that generate automatic decisions, learning from successes and errors.

But if there is something that is causing the entire logistics process to change in a radical way, it is robotization. Until date, a large number of personnel dedicated to logistics processes such as product and merchandise handling, order issuance, warehouse and inventory control or replenishment management were needed. However, robots are replacing these functions and ending, to a large extent, with the need for labour. These are capable of carrying a load of up to 500 Kg from one end to another of the warehouse. And even move it from one warehouse to another. In the same way, they can rotate 360 ​​degrees on its axis, rise to load merchandise or deposit it gently at any point. We need to consider that they do much faster than humans without getting sick or requiring rest.

Within five minutes, the robots recharge and have a range of 4 or 5 hours. So they can cover a full 8-hour shift with a total of ten minutes of recharging. Consequently, they can perfectly cover three daily shifts. It has no conflict between them or collective claims and they are immediately replaceable in case of breakdown or need for maintenance.

The logistics robotization process is generating profound changes in the business model that affect all areas, especially the human resource management.

According to reports from the World Economic Forum, by 2025 the replacement of human personnel with robots in all basic professional areas will have reached 52%. This means the loss of countless unskilled jobs. That will be compensated with the creation of 58 million qualified jobs, necessary for the robotic revolution, in the next 10 years.

It is not difficult to think that technical qualifications will be one of the challenges to overcome in this whole process.

Featured examples of robotization in large multinationals

Two of the most prominent precursors within this logistics reorganization have been two giants of online commerce; Alibaba and Amazon.

Amazon’s experience

Amazon has more than one hundred thousand robots dedicated to managing the orders of its customers and all its stores are currently automated. Quite the opposite of destroying employment, the company has doubled the workforce since 2016, currently having more than 500,000 workers.

Kiva robots, used by the firm, can easily replace physical work and repetitive activities that are easily programmable. But the same does not happen with another set of required skills that are demanded in new positions to add value.

Alibaba’s model

With its logistics model, it has facilitated the penetration of thousands of companies in different international markets. Process automation and artificial intelligence are the engines of productivity in our day and this, in turn, is the key factor in competitiveness.

Only through these logistic processes is possible to manage the huge volume of orders for days like Black Friday or the Alibaba shopping festival.

JD’s case

Another Chinese marketing giant, JD has recently surpassed Alibaba with a warehouse capable of processing more than 200,000 orders daily with only the supervision of 4 people. The objective of this company is to provide service to all of China on the same day as long as the order is generated before 11 am.

The company has not only invested millions in robots for warehouses but also done so in the incorporation of automatic systems in trucks, means of transport and distribution drones.

Prof. Raul V. Rodriguez

Machine Learning

The Future of the Maritime Logistics Industry: Unmanned ships from 2020

There are no drones only across the sky but also on land and sea and Rolls Royce has focused on the latter for its commercial strategy as far as vessels are concerned.

The company, which no longer manufactures cars -transferred the automobile division to BMW- is a conglomerate that operates in the aeronautical, aerospace, maritime and energy sectors. They have a clear-cut commitment to the seas: launch unmanned ships by mid-2020.

In the meantime, the Rolls-Royce Blue Ocean research team has already launched a virtual reality prototype in its office in Alesund, Norway, which simulates the views from a ship’s command bridge in 360 degrees. The manufacturer hopes that ship captains can maneuver hundreds of unmanned ships from the ground, without any need to approach the sea.

The idea is that during this year the first fleet of unmanned ships will be built. The first would be tugboats or ferries, boats that make simple, short-sized journeys in controlled environments. At first, all risks must be minimized, in order to avoid any possibility of unforeseen events.

The next stage would be the launch of cargo ships, with increasing complexity, especially because they sail in international waters. As of today, there is no legislation that covers unmanned commercial shipping. And the approval of international regulation is always slower than that processed by individual countries.

Unmanned ships, according to Rolls Royce, will reduce operating costs by 20%. Companies, therefore, buy ships to increase their profit margins. The other side of technology is the possible loss of jobs. It will not be necessary to have a crew either a large contingent of security personnel. However, piracy will surely remain a threat that requires the presence of minimal security personnel while keeping in mind that there will not be as many lives at stake in the absence of crew members as the risk for cargo theft.

Although, Rolls-Royce pointed out that new jobs will be created. The operations will have to be performed from the ground. It is an unmanned craft, not autonomous. Cybersecurity will be a key element assuring secured communications links between the ship and land, hence new profiles will be necessary.

By replacing the control bridge along with the other systems where the crew is usually accommodated – including electricity, air conditioning, water, and waste treatment system- the ships will withstand more cargo, reducing costs and increasing revenue. In addition to this, according to the initial calculations, these ships will be 5% lighter and consume between 12 to 15% less fuel ensuring a greener performance. Similarly, electric fuel-free ships are being researched in order to consider their implementation.

Waterfall Office Park Gardens Elevation, Midrand. 1686 | +27 10 634 0880 |

Machine Learning Africa provides insight into emerging machine learning technologies and their inevitable impact in transforming Africa. Provides a platform where innovators, technology vendors, end users and enthusiasts discuss latest innovations and technologies that transform businesses and the broader society.

Contact Us

Waterfall Office Park Gardens Elevation, Midrand. 1686
Tel: +27 10 634 0880


Machine Learning

Will Artificial Intelligence reach the level of the human intellect by 2040?

Technological singularity is a hypothesis that predicts that there will come a time when artificial intelligence will be able to improve itself recursively. In theory, machines that are capable of creating other machines even more intelligent, resulting in intelligence far superior to human beings and, which could be even more shocking, beyond our control.

AI, Machine Learning, Neural Networks… these are terms that transmit feelings which are equally of hope and fear of the unknown.

In the next 20 years, there will be more technological changes than in the last 2 millennia. The technology is much faster than the brain – a calculator multiplies 5-digit numbers in tenths of a second – but it works differently, for example, it does not have the level of connections equivalent to that of neurons in a human brain.

However, if the exponential speed of Moore’s law does not stop and the investigations of neural networks of giant corporations such as Google continue to advance by 2040 the degree of technological integration in our lives will far exceed the capacity of the human brain.

The word singularity was taken from astrophysics: a point in space-time – for example, inside a black hole – in which the rules of ordinary physics are not lost. It was associated with the explosion of artificial intelligence during the 1980s by science-fiction novelist Vernor Vinge. At a NASA symposium in 1993, Vinge predicted that in 30 years there would be technological means to create superhuman intelligence called Singleton which refers to a “world order in which there is a single decision-making entity at the highest level, capable of exerting effective control over its domain and preventing internal or external threats to its supremacy”. In addition to this, he assured that, shortly after, we would reach the end of the human era.

Throughout history, some technological advances have caused fear. The fear of the new and the unknown is understandable, however, all technologies can be modified for good or for evil, as you can use fire to heat and cook food, or to burn people

In the case of the singularity, it seems clear that one must be cautious, regulating its development but without limiting it and, above all, trying to ensure that this future artificial intelligence learns from ethical and moral values, as well as from mistakes and successes of the species. We must be clear in our conception of the term. Human beings and machines are meant to co-exist in symbiosis and not rivalry. 

Mortality as an “option” by 2045?

On the other hand, we could analyze if mortality will be “optional” by 2045. Google has already started extravagant research initiatives as they realized that curing aging is possible and that is why they are creating companies such as ‘Calico’ or ‘Human Longevity’, which are investigating it, but also non-profit organizations such as the Methuselah Foundation. It is evident that the possibilities are real since immortality already exists in nature. Some cells are immortal and the stem cells affected the quality of reproducing indefinitely, just like cancer cells.

One of the steps to achieve this is to fully comprehend the structure of incurable diseases today, and then eradicate them. Thus, as it happens with HIV, a controllable chronic disease, or diabetes. We must propose the same with aging: turn it into a controllable chronic disease, and later on, cure it for good. It is essential to begin human trials with rejuvenation technologies that have been shown useful in other animals leading to advancements in human clinical trials as well. 

Prof. Raul V. Rodriguez is an Asst. Professor at Universal Business School.

Machine Learning

Predicting people’s driving personalities

Self-driving cars are coming. But for all their fancy sensors and intricate data-crunching abilities, even the most cutting-edge cars lack something that (almost) every 16-year-old with a learner’s permit has: social awareness.

While autonomous technologies have improved substantially, they still ultimately view the drivers around them as obstacles made up of ones and zeros, rather than human beings with specific intentions, motivations, and personalities.

But recently a team led by researchers at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) has been exploring whether self-driving cars can be programmed to classify the social personalities of other drivers, so that they can better predict what different cars will do — and, therefore, be able to drive more safely among them.

In a new paper, the scientists integrated tools from social psychology to classify driving behavior with respect to how selfish or selfless a particular driver is.

Specifically, they used something called social value orientation (SVO), which represents the degree to which someone is selfish (“egoistic”) versus altruistic or cooperative (“prosocial”). The system then estimates drivers’ SVOs to create real-time driving trajectories for self-driving cars.

Testing their algorithm on the tasks of merging lanes and making unprotected left turns, the team showed that they could better predict the behavior of other cars by a factor of 25 percent. For example, in the left-turn simulations their car knew to wait when the approaching car had a more egoistic driver, and to then make the turn when the other car was more prosocial.

While not yet robust enough to be implemented on real roads, the system could have some intriguing use cases, and not just for the cars that drive themselves. Say you’re a human driving along and a car suddenly enters your blind spot — the system could give you a warning in the rear-view mirror that the car has an aggressive driver, allowing you to adjust accordingly. It could also allow self-driving cars to actually learn to exhibit more human-like behavior that will be easier for human drivers to understand.

“Working with and around humans means figuring out their intentions to better understand their behavior,” says graduate student Wilko Schwarting, who was lead author on the new paper that will be published this week in the latest issue of the Proceedings of the National Academy of Sciences. “People’s tendencies to be collaborative or competitive often spills over into how they behave as drivers. In this paper, we sought to understand if this was something we could actually quantify.”

Schwarting’s co-authors include MIT professors Sertac Karaman and Daniela Rus, as well as research scientist Alyssa Pierson and former CSAIL postdoc Javier Alonso-Mora.

A central issue with today’s self-driving cars is that they’re programmed to assume that all humans act the same way. This means that, among other things, they’re quite conservative in their decision-making at four-way stops and other intersections.

While this caution reduces the chance of fatal accidents, it also creates bottlenecks that can be frustrating for other drivers, not to mention hard for them to understand. (This may be why the majority of traffic incidents have involved getting rear-ended by impatient drivers.)

“Creating more human-like behavior in autonomous vehicles (AVs) is fundamental for the safety of passengers and surrounding vehicles, since behaving in a predictable manner enables humans to understand and appropriately respond to the AV’s actions,” says Schwarting.

To try to expand the car’s social awareness, the CSAIL team combined methods from social psychology with game theory, a theoretical framework for conceiving social situations among competing players.

The team modeled road scenarios where each driver tried to maximize their own utility and analyzed their “best responses” given the decisions of all other agents. Based on that small snippet of motion from other cars, the team’s algorithm could then predict the surrounding cars’ behavior as cooperative, altruistic, or egoistic — grouping the first two as “prosocial.” People’s scores for these qualities rest on a continuum with respect to how much a person demonstrates care for themselves versus care for others.

In the merging and left-turn scenarios, the two outcome options were to either let somebody merge into your lane (“prosocial”) or not (“egoistic”). The team’s results showed that, not surprisingly, merging cars are deemed more competitive than non-merging cars.

The system was trained to try to better understand when it’s appropriate to exhibit different behaviors. For example, even the most deferential of human drivers knows that certain types of actions — like making a lane change in heavy traffic — require a moment of being more assertive and decisive.

For the next phase of the research, the team plans to work to apply their model to pedestrians, bicycles, and other agents in driving environments. In addition, they will be investigating other robotic systems acting among humans, such as household robots, and integrating SVO into their prediction and decision-making algorithms. Pierson says that the ability to estimate SVO distributions directly from observed motion, instead of in laboratory conditions, will be important for fields far beyond autonomous driving.

“By modeling driving personalities and incorporating the models mathematically using the SVO in the decision-making module of a robot car, this work opens the door to safer and more seamless road-sharing between human-driven and robot-driven cars,” says Rus.

The research was supported by the Toyota Research Institute for the MIT team. The Netherlands Organization for Scientific Research provided support for the specific participation of Mora.

Machine Learning

ML Africa successfully hosted the inaugural AI & The Future of Healthcare Summit

Artificial intelligence and machine learning are the most trending and dominating technologies of our times. These are shaping the future and impacting on our daily lives. For businesses and government, adoption and agile adoption of these technologies is imperative.

Machine learning Africa celebrates its successful hosting of the inaugural AI & The Future of Healthcare Summit at Hilton Sandton on the 30th of October 2019. It was a wonderful event where technology enthusiasts were sharing insights into the development of AI driven healthcare solutions that improve patient outcomes.

The discussions focused on the future of healthcare, patient engagement, the public and private sector collaboration, digital health strategy, AI in Radiology, precision medicine, the future of robots in healthcare, diagnostic technologies and upskilling healthcare workforce.

Key note speakers included: Prof. Nelishia Pillay, Head of the Department of Computer Science at the University of Pretoria, Johan Steyn, AI Enthusiast, Portfolio Lead: DevOps & Software at IQBusiness South Africa, Joel Ugborogho, Founder of CenHealth, Dr. Jonathan Louw, MB.ChB, MBA, CEO of the South African National Blood Service (SANBS), Basia Nasiorowska, CEO at NEOVRAR, Josh Lasker, Co-Founder, Abby Health Stations, Dr. Jaishree Naidoo, Paediatric Radiologist and CEO of Envisionit Deep AI, Prof. Antonie van Rensburg, PrEng, Chief Digital Officer IoTDot4, Dr. Darlington Mapiye (PhD) Technical Lead for the data driven healthcare team at IBM Research Africa, Dr. Boitumelo Semete, Executive Cluster Manager, CSIR, and  Yusuf Mahomedy, Chief Executive of the Association Executive Network of Southern Africa (AENSA)

The event was made possible through partnership with Envisionit Deep AI, a medical technology company that utilizes artificial intelligence to streamline and improve medical imaging diagnosis for radiologists. Their AI model RADIFY will augment and improve the radiology reading and thereby relieve the bottlenecks we face in medical imaging. Other event partners present were Evolutio, SANBS, IQBusiness, IoTDot4 and ICITP.

 If you would like to increase your proficiency further in emerging technologies and deploy the most effective strategies within your organization, the Digital Health workshop would be another exciting and relevant event to consider. Entitled ‘Accelerating Digital Health Services’, the workshop is brought in partnership with Cenhealth, on the 5th of December 2019. In preparation for the upcoming changes in the healthcare industry, it is imperative for all healthcare institutions not be left behind in their digital transformation journey.