Connect with us

Featured

IBM aims AI at Africa

Published

on

Artificial intelligence is seen as an advanced and dangerous technology, but IBM is showing how it can help solve problems for Africa, writes ARTHUR GOLDSTUCK.

Artificial intelligence has been getting a bad rap lately. It’s been blamed for the way Facebook and Twitter promote fake news on the basis of showing people more of what they are already click through to read. That in turn is credited by some with handing Donald Trump the US presidency. It’s also been blamed for introducing bias into anything from home loan approvals to facial recognition systems, which tend to be more accurate with Caucasian faces than those of others.

But a team of scientists at a research facility in Braamfontein, Johannesburg, hopes to change all that. 

“We are working on beneficial AI,” says Tapiwa  Chiwewe, a scientist at IBM Research, which opened its second global research lab in Africa at the Tshimologong precinct in 2016. “We want to create AI that benefits people. In Africa, there are all kinds of socio-economic problems and we’re in a unique position to create AI that improves people’s lives. But we also want to create economic and social value.”

Screen Shot 2018-04-17 at 9.35.28 PM

The emphasis is not on AI for its own sake, or operating independently of humans, but on creating symbiotic systems that help humans perform their roles better.

“These range from data driven healthcare to digital urban ecosystems,” says Craig Holmes, vice president for Industry Solutions and Business Development at IBM Middle East and Africa. “We’re researching air-quality management in partnership with the city of Johannesburg and Tshwane, we’re looking at traffic management in cities, and even wildfire risk prediction in the Cape Town area.”

Many of these projects will deliver benefits in the medium- to long-term future, but at least one is ready to roll out. It will be launched this Sunday, April 22, marking Earth Day. Appropriately, it is aimed at promoting sustainable energy.

Screen Shot 2018-04-17 at 9.35.37 PM

It’s called EmpowerSolar, and is a web-based application for designing solar powered solutions for domestic and business use. 

IBM explains on the site: “The app uses a sophisticated algorithm to determine an optimal solution, along with an estimated price based on current typical prices. The algorithm takes into account the amount of sunlight at your chosen location, along with the specified usage of the various electrical appliances chosen.”

Screen Shot 2018-04-17 at 9.36.02 PM

The app was developed by Dr Ashley Gritzman, a young scientist at IBM Research. By coincidence, he had graduated as an electrical engineer and was an installer of solar systems before joining IBM. 

“I had a small startup, assessing people’s energy usage, monitoring household or small business consumption, and told them how to reduce their energy use, ranging from changing old light bulbs to LED, to using solar geysers for water heating. Then we’d spec the systems, and design and install. I saw there was a big divide between home owners who may want energy efficiency, and the knowledge of installers and retailers of the technology.

Screen Shot 2018-04-17 at 9.35.53 PM

“When I joined this lab, they asked, what contribution can we make to the energy question in South Africa? I saw a big gap of information. For example, solar is regarded as way too expensive for many and they wouldn’t look at it. But in fact, people have no idea what it will cost. If they assume it will cost hundreds of thousands of rands, they won’t even pick up the phone to enquire. I wanted to bridge the gap between homeowners and new technology.”

Part of Gritzman’s motivation was the dramatic fall in solar costs: the price of panels has fallen 50% in the last 5 years. The Renewable Energy Independent Power Producer Procurement (REIPPP) programme, designed to encourage private investment in renewable energy, has seen contract prices fall from R4 to 65c per Kilowatt Hour since 2011.

“South Africa has great resources in both solar and wind, and there is no longer a question that it is cheaper than new-built coal or nuclear. But it has to be super user-friendly, so that my parents can use it; it has to be technically accurate and pricing has to be correct, or else it’s not useful. And the electrical designs it comes up with have to work and be correct.

“When I was doing the solar stuff myself, a design would take a lot of time and a lot of calculations to do an initial quote. Then they also want to run the garage lights, for example, and you have to redo it. The app will be useful to home owners, but also to installers to do a quick design and costing.”

Gritzman’s PhD was in machine learning and image processing, which fits in perfectly with the artificial intelligence focus of the Research Lab and IBM’s Watson AI engine, which has been used for numerous medical breakthroughs, ranging from cancer research to tuberculosis treatment. However, EmpowerSolar fed a personal passion, and he was quickly able to drive it from idea to reality.

Screen Shot 2018-04-17 at 9.35.22 PM

“It’s a responsive web site, it looks like an app, but doesn’t have to be downloaded. You choose your location, and we ask a few questions that are easy to understand. The big difference is in how much sun you get.  We’ve got little sliders, and ask people to use those to select direction, angle of roof and appliances they’d like to use with solar energy.

“When I was an installer, we’d just say how many kilowatt hours do you use a month, and most people don’t know. So we build this from the ground up, with categories, like home office, kitchen, and lighting, and within each we have a database of generic appliances.”

Part of the purpose of the service is to promote energy-efficient behaviour. As users indicate the appliances they use, messages pop up that, for example, advise that an electric geyser uses a large amount of electricity, and indicates the cost and benefit of installing gas or solar. Nothing is hidden. The application includes costs for solar panels, batteries, charge controllers, inverters and installation costs.

Gritzman is as enthusiastic about South Africa’s advantages as he is about solar power.

“South Africa really is sunny. If you compare us with Germany, one of the world leaders in solar, the least sunny spot in South Africa still has more sun than the most sunny city in Germany. That puts in perspective what opportunity we have.”

* Try out the EmpowerSolar application at https://empowersolar.mybluemix.net/

Featured

IoT at tipping point

We have long been in the hype phase of IoT, but it is finally taking on a more concrete form illustrating its benefits to business and the public at large, says PAUL RUINAARD, Country Manager at Nutanix Sub-Saharan Africa.

Published

on

People have become comfortable with talking to their smartphones and tasking these mini-computers to find the closest restaurants, schedule appointments, and even switch on their connected washing machines while they are stuck in traffic.

This is considerable progress from those expensive (and dated) robotic vacuum cleaners that drew some interest a few years ago. Yes, being able to automate cleaning the carpets held promise, but the reality failed to deliver on those expectations.

However, people’s growing comfort when it comes to talking to machines and letting them complete menial tasks is not what the long-anticipated Internet of Things (IoT) is about. It really entails taking connectedness a step further by getting machines to talk to one another in an increasingly digital world filled with smart cities, devices, and ways of doing things.

We have long been in the hype phase of IoT, but it is finally taking on a more concrete form illustrating its benefits to business and the public at large. The GSM Association predicts that Africa will account for nearly 60 percent of the anticipated 30 billion connected IoT devices by 2020.

Use cases across the continent hold much promise. In agriculture, for example, placing sensors in soil enable farmers to track acidity levels, temperature, and other variables to assist in improving crop yields. In some hotels, infrared sensors are being used to detect body heat so cleaning staff now when they can enter a room. In South Africa, connected cars (think telematics) are nothing new. Many local insurers use the data generated to reward good driver behaviour and penalise bad ones with higher premiums.

Data management

The proliferation of IoT also means huge opportunity for businesses. According to the IDC, the market opportunity for IoT in South Africa will grow to $1.7 billion by 2021. And with research from Statista showing that retail IoT spending in the country is expected to grow to $60 million by the end of this year (compared to the $41 million of 2016), there is significant potential for connected devices once organisations start to unlock the value of the data being generated.

But before we get a real sense of what our newly-connected world will look like and the full picture of the business opportunities IoT will create, we need to put the right resources in place to manage it. With IoT comes data, more than we can realistically imagine, and we are already creating more data than ever before.

Processing data is something usually left to ‘the IT person’. However, if business leaders want to join the IoT game, then it is something they must start thinking about. Sure, there are several ways to process data but they all link back to a data centre, that room or piece of equipment in the office, or the public data centre down the road. Most know it is there but little else, other than it has something to do with data and computers.

Data centres are the less interesting but very essential tools in all things technology. They run the show, and without them we would not be able to do something as simple as send an email, let alone create an intricate system of connected devices that constantly communicate with each other.

Traditionally, data centres have been large, expensive and clunky machines. But like everything in technology, they have been modernised over the years and have become smaller, more powerful, and more practical for the digital demands of today.

Computing on the edge

Imagine real-time face scanning being used at the Currie Cup final or the Chiefs and Pirates derby. Just imagine more than a thousand cameras in action, working in real time scanning tens of thousands of faces from different angles, creating data all along the way and integrating with other technology such as police radios and in-stadium services.

As South Africans, we know all too well that the bandwidth to process such a large amount of data through traditional networks is simply not good enough to work efficiently. And while it can be run through a large core or public data centre, the likelihood of one of those being close to the stadium is minimal. Delays, or ‘latency and lag time’, are not an option in this scenario; it must work in real time or not at all.

So, what can be done? The answer lies in edge computing. This is where computing is brought closer to the devices being used. The edge refers to devices that communicate with each other. Think of all those connected things the IoT has become known for: things like mobile devices, sensors, fitness trackers, laptops, and so on. Essentially anything that is ‘remote’ that links to the Web or other devices falls under this umbrella. For the most part, edge computing refers to smaller data centres (those in the edge) that can process the data required for things like large-scale facial recognition.

At some point in the future, there could be an edge data centre at Newlands or The Calabash that processes the data in real time. It would, of course, also be connected to other resources such as a public or private cloud environment, but the ‘heavy lifting’ is done where the action is taking place.

Unfortunately, there are not enough of these edge resources in place to match our grand IoT ambitions. Clearly, this must change if we are to continue much further down the IoT path.

Admittedly, edge computing is not the most exciting part of the IoT revolution, but it is perhaps the most necessary component of it if there is to be a revolution at all.

Continue Reading

Featured

Don’t panic! Future of work is still human

Published

on

The digital age, and the new technologies it’s brought with it – blockchain, artificial intelligence (AI), robotics, augmented reality and virtual reality – is seen by many as a threat to our way of life as we know it. What if my job gets automated? How will I stay relevant? How do we adapt to the need for new skills to manage customer expectations and the flood of data that’s washing over us?

The bad news is that the nature of work has already changed irrevocably. Everything that can be automated, will be. We already live in an age of “robot restaurants”, where you order on a touch screen, and machines cook and serve your food. Did you notice the difference? AmazonGo is providing shopping without checkout lines. In the US alone, there are an estimated 3.4 million drivers that could be replaced by self-driving vehicles in 10 years, including truck drivers, taxi drivers and bus drivers.

We’re not immune from this phenomenon in Africa. In fact, the World Economic Forum (WEF) predicts that 41% of all work activities in South Africa are susceptible to automation, compared to 44% in Ethiopia, 46% in Nigeria and 52% in Kenya. This doesn’t mean millions of jobs on the continent will be automated overnight, but it’s a clear indicator of the future direction we’re taking.

The good news is that we don’t need to panic. What’s important for us in South Africa, and the continent, is to realise that there is plenty of work that only humans can do. This is particularly relevant to the African context, as the working-age population rises to 600 million in 2030 from 370 million in 2010. We have a groundswell of young people who need jobs – and the digital age has the ability to provide them, if we start working now.

Make no mistake, there’s no doubt that this so-called “Fourth Industrial Revolution” is going to disrupt many occupations. This is perfectly natural: every Industrial Revolution has made some jobs redundant. At the same time, these Revolutions have created vast new opportunities that have taken us forward exponentially.

Between 2012 and 2017, for example, it’s estimated that the demand for data analysts globally grew by 372%, and the demand for data visualisation skills by more than 2000%. As businesses, this means we have to not only create new jobs in areas like data science and analytics, but reskill our existing workforces to deal with the digital revolution and its new demands.

So, while bus drivers and data clerks are looking over their shoulders nervously right now, we’re seeing a vast range of new jobs being created in fields such as STEM (Science, Technology, Engineering and Mathematics), data analysis, computer science and engineering.

This is a challenge for Sub-Saharan Africa, where our levels of STEM education are still not where they should be. That doesn’t mean there are no opportunities to be had. In the region, for example, we have a real opportunity to create a new generation of home-grown African digital creators, designers and makers, not just “digital deliverers”. People who understand African nuances and stories, and who not only speak local languages, but are fluent in digital.

This ability to bridge the digital and physical worlds, as it were, will be the new gold for Africa. We need more business operations data analysts, who combine deep knowledge of their industry with the latest analytical tools to adapt business strategies. There will also be more demand for user interface experts, who can facilitate seamless human-machine interaction.

Of course, in the longer term, we in Africa are going to have to make some fundamental decisions about how we educate people if we’re going to be a part of this brave new world. Governments, big business and civil society will all have roles to play in creating more future-ready education systems, including expanded access to early-childhood education, more skilled teachers, investments in digital fluency and ICT literacy skills, and providing robust technical and vocational education and training (TVET). This will take significant intent not only from a policy point of view, but also the financial means to fund this.

None of this will happen overnight. So what can we, as individuals and businesspeople, do in the meantime? A good start would be to realise that the old models of learning and work are broken. Jenny Dearborn, SAP’s Global Head of Learning, talks about how the old approach to learning and work was generally a three-stage life that consisted largely of learn-work-retire.

Today, we live in what Ms Dearborn calls the multi-stage life, which includes numerous phases of learn-work-change-learn-work. And where before, the learning was often by rote, because information was finite, learning now is all about critical thinking, complex problem-solving, creativity and innovation and even the ability to un-learn what you have learned before.

Helping instill this culture of lifelong learning, including the provision of adult training and upskilling infrastructure, is something that all companies can do, starting now. The research is clear: even if jobs are stable or growing, they are going through major changes to their skills profile. WEF’s Future of Jobs analysis found that, in South Africa alone, 39% of core skills required across all occupations will be different by 2020 compared to what was needed to perform those roles in 2015.

This is a huge wake-up call to companies to invest meaningfully in on-the-job training to keep their people – and themselves – relevant in this new digital age. There’s no doubt that more learning will need to take place in the workplace, and greater private sector involvement is needed. As employers, we have to start working closely with should therefore offer schools, universities and even non-formal education to provide learning opportunities to our workers.

We can also drive a far stronger focus on the so-called “soft skills”, which is often used as a slightly dismissive term in the workplace. The core skills needed in today’s workplace are active listening, speaking, and critical thinking. A quick look at the WEF’s “21st Century Skills Required For The Future Of Work” chart bears this out: as much as we need literacy, numeracy and IT skills to make sense of the modern world of work, we also need innately human skills like communication and collaboration. The good news is that not only can these be taught – but they can be taught within the work environment.

It sounds almost counter-intuitive, but to be successful in the Digital Age, businesses are going to have to go back to what has always made them strong: their people. Everyone can buy AI, build data warehouses, and automate every process in sight. The companies that will stand out will be those that that focus on the things that can’t be duplicated by AI or machine learning – uniquely human skills.

I have no doubt that the future will not be humans OR robots: it will be humans AND robots, working side by side. For us, as businesspeople and children of the African continent, we’re on the brink of a major opportunity. We just have to grasp it.

Continue Reading

Trending

Copyright © 2018 World Wide Worx