The four teenaged girls who ran away from the St Dominic’s Children’s Home in Belmont last Wednesday have all been found.
You are here
Hosein leads T&T's think tank with TTLAB
The term “think tank” describes an institution which pays people to think, create, and pursue social and scientific research for their own sake. It has been a largely metropolitan phenomenon till now, but has finally reached the Caribbean outside the academe in TTLAB, thanks to Prof Patrick Hosein, a MIT-trained computer scientist who works at the University of the West Indies (UWI). He is also the holder of several patents in cellular phone and Internet technology, and is the ANSA Caribbean Awards Science & Technology Laureate 2015.
One of Prof Hosein’s peeves about regional academia and development agendas is that there has been much talk but little real understanding about innovation and its requirements. What we think is innovation, he says, is usually just importing and modifying foreign technology. As a result, many opportunities for innovation are lost, or simply wither on the local vine. One example which was crucial to TTLAB involved a student at UWI, St Augustine, where Prof Hosein teaches computer science.
The UWI student, Stefan Hosein (no relation to Prof Hosein), had been to Nasa (National Aeronautics and Space Administration in the US) on an internship/scholarship, thanks to the T&T National Institute of Higher Education, Research, Science & Technology (Niherst). One of the conditions of the scholarship was that he had to return home to serve in a government institution.
When he returned to the UWI, said Prof Hosein, he was given relatively menial tasks to perform, like web development. He was capable of much more, and Prof Hosein decided to engage in an experiment: pay him a stipend, allow him to pursue research, and see if he could come up with a publishable paper within three months.
Stefan went on to publish four papers in peer-reviewed international conferences during a one-year period. These were based on his research at TTLAB but he also collaborated with Nasa on other publications during this period. This was TTLAB’s first fellow. “TTLAB has helped me tremendously,” Stefan said.
“I was able to produce research (something that is rare for a person without a postgraduate degree in Trinidad) which helped me to secure acceptance to The University of Cambridge for my postgraduate degree. Additionally, I was able to attend one of the premier conferences in my research area, this allowed me to network with those who are top in the field and get a better understanding of what the state-of-the-art research looks like.”
TTLAB is funded entirely by Prof Hosein’s company, the T&T Network Information Centre (TTNIC). It seeks and attracts fellows who range from undergraduates pursuing higher degrees to scientists who cannot do research in their workplaces, offers them funding, and lets them pursue projects they want to. The results have been spectacular. From its one student/researcher, a little over a year ago the project now has 19 fellows and researchers. One is another Nasa intern, Inzamam Rahaman, who was introduced to Prof Hosein by his first protege.
Stefan’s work is mostly in Machine Learning, but Rahaman’s work is in mathematical optimisation, directed at using social networks, like Twitter and Facebook. Additionally, he is pursuing research (with Shanta Sukhoo) on “how we can characterise the relationship between collaboration and innovation en route to building recommendation systems for team building”. He is also interested in the Internet of Things and its potential in development.
These are just two researchers and fields, there are many more. On its website (http://lab.tt) TTLAB lists the variety of its achievements: 15 publications and conference papers, which is remarkable by any standard for a year’s endeavour. The topics addressed range from optimising electricity grids, gauging emotional responses from social media (“Sentiment Classification from Twitter Feeds”), an app for the detection of dangerous driving, and the genetics of non-communicable diseases in the region.
To spur its researchers, any member of TTLAB, who has an accepted conference paper, is provided with a travel grant to attend and present in at least one conference per year. TTNIC also funds full fellowships for two PhD and two MPhil students. This year, three full scholarships have been awarded to three 2016 first-class honours students to pursue their MSc degree in Computer Science at the UWI.
Away from its research concerns is a cultural one: TTLAB is also a concrete response to one of Prof Hosein’s bugbears: that technology and expertise are imported when they are available at the same quality for a cheaper price locally. This was also a constant complaint of the late Prof Dave Chadee, a world famous entomologist, contracted by international agencies, who was all but ignored in T&T.
“A lot of local companies use Huawei technology,” said Prof Hosein. “I worked at Huawei and have about 15 patents in my name which are integrated into their products. So my work is being used, but no one will accept my help when I offer it locally.”
TTNIC manages the registration of “.tt” domain names. It also provides free domain names (under edu.tt) to all educational institutions and offers free website hosting to those that are registered with the Government. In addition, it provides sponsorships to many local ICT related events.
Domain name policies are decided by a multi-stakeholder advisory group, the TTMAG (see mag.tt). TTNIC is also active in Open Data and manages the sites data.tt and maps.tt. Its funds are limited, but being entirely self-funded means it doesn’t have to answer to impatient businesses or wait on bureaucratic funders.
And here arises a question of why TTLAB’s establishment had to take so long, and await a single individual to do it. T&T has three universities and (up till 2016) universal free tertiary education, yet, there has been no comparable endeavour.
Cultural problem hindering research and development
Incoming UWI, St Augustine, Pro Vice Chancellor Prof Brian Copeland disagrees about the disparity in output. There has been much research and innovation at the UWI, he said.
“UWI has a range of units and centres that conduct research on a variety of topics. The Seismic Unit, for example. Attitudes to the full range of the S&T spectrum (blue sky to applied) have been changing over the past few years. The work of St Clair King’s Real Time Systems Group, the work of those who generated patents over the years (Prof Ramsey Sanders and others, including myself), the work at the Cocoa Research Centre which has just seen the signing of MOUs with external commercial entities, and the work at the Steelpan Initiatives Project for developing the GPan and PHI for commercial production.”
As to output and quality of work, said Prof Copeland: ”My most recent experience in Engineering was of students with modest potential, as measured by the coursework GPA, rising to the challenge of final year projects and, in some cases, submitting award-wining papers internationally.” The UWI, he continues, has also designed apps for local fisherfolk, its Medical Sciences staff is prolific in their work, and its Dean is ranked in the top 100 “most influential papers on Chronic Obstructive Pulmonary Disease (COPD).”
There is no doubt the UWI has had a strong impact regionally and produces high-quality research. But two complaints about the UWI are its disconnectedness from the societies it serves, and its internal politics. Prof Copeland acknowledges this. Prof Copeland said: “The conditions at UWI are indeed less than ideal and the staff, who have rallied, despite this, are to be celebrated. Patrick is an absolute gem primarily because of his experience at Bell Labs (we do not get many of those) and his prolific patent [record].
“He has attracted students to TTLAB because of this. He has seen the mechanism of concept to commercial reality first hand at Bell. It costs to get people to that level of achievement…you should ask him to estimate the cost to Bell Labs for his patents. This country grudgingly invests 0.04 per cent of its GDP in RDI.
The most direct RDI subvention by government of Trinidad & Tobago is a paltry $5 million a year. Industry and commerce provide much less than that. The one or two genuine and valid attempts to produce new products with international market potential have been all but slaughtered. There is a cultural problem in the way of progress.”
All this notwithstanding, Joycelyn Lee Young, a former acting president of Niherst, said TTLAB’s “R&D productivity is phenomenal. The Real Time Systems Group run by Prof St Clair King in the 1990s was close but I don’t recall that it achieved this much in a one-year timeframe. TTLAB would have achieved a first to my mind with Stefan Hosein doing four papers at conferences and Imzaman Rahaman having three, both with just a BSc.”
An important component of TTLAB’s success, continued Ms Lee Young, is its personality, and the personality of its founder. She said: “I am amazed to see how Patrick/TTLAB took two of our Nasa interns to new heights in R&D—from the conceptualisation stage to execution, the robust methodology, and high-level problem solving undertaken. Both interns had just a BSc but their work was at the level of post-grad students at MIT or Stanford. To me it showed what a mentor who has worked on the world stage could do with bright young minds in a local setting. He is passionate about developing young researchers, inspiring them to believe in themselves and guiding them towards achieving high quality R&D.”
Apropos of Prof Copeland’s identification of the “cultural problem” hindering research and development, an important part of TTLAB’s character is its upending of conventional development approaches. “There’s a lot of emphasis locally on building large buildings, and then there’s no resources left for research,” says Prof Hosein. TTLAB is the opposite. It has no headquarters or bureaucracy.
The company simply selects the best students (or invites applicants it believes are gifted) gives them money, and lets them do what they want. He treats them the way that he likes to be treated which is to be given autonomy to explore and have complete time flexibility (they can work from home if they wish and, like him, many prefer working at nights). Unfortunately he cannot pay as much as he thinks they deserve to be paid but the Fellows actually receive a larger stipend than that provided by UWI scholarships.
Lee Young agrees: “In Patrick’s field T&T can become a global player in R&D—T&T does not need expensive huge labs. We are dealing with brains—reason, logic, computation, and software skills. We can develop and harvest this resource in a short space of time—compared to R&D in other disciplines.”
This approach is particularly useful, says Prof Hosein, to those scientists and engineers who work in day jobs, but are not allowed to do the kind of research they want to do. With a little extra money, they can take time off from work, and spend it on the things that matter more. He believes that a “reverse-intern” program whereby bright engineers/scientists in industry get to perform research under the supervision of UWI Faculty might be feasible and all three parties will benefit.
Indeed, he says, it was his prize from the ANSA Caribbean Awards that allowed him to take time off from his teaching at UWI to pursue his own research and help fund his first experimental student.
Another innovative practice is encouraging students to think of what they would like to pursue, as well as encouraging topics of local use and interest but with global scope. “To publish internationally, you have to be more broadly focused than just Trinidad & Tobago,” says Prof Hosein, “but there are things here that qualify.” One of these things is electricity grid efficiency.
TTLAB (through a RDI Smart Grid project lead by Dr Sharma at UWI) approached T&TEC for data from their meters which they analysed and developed an algorithm and an application which could provide significant savings if applied since comparable foreign software is expensive. The project, unfortunately, has been frozen.
But engineering is not the only area that can benefit. The UWI (and every university library) maintains subscriptions to international journals which, must be paid for in US currency. Analysing the data on how UWI members access the journals, TTLAB derived an optimization algorithm which recommends which journals should be subscribed to, and which subscriptions can be halted, and individual articles purchased instead. This would mean no loss of access for students and faculty, but millions of dollars can potentially be saved by the university.
The biggest problem in putting this knowledge to work, says Prof Hosein is local inertia. “We have tried to talk to local companies, but they only see us as app developers, and anyone with computer knowledge can develop an app. Most don’t see the value in our expertise in Machine Learning, Optimization, Decision Support Systems and so on. Just as in the above cases we can use this expertise to help them achieve significant savings.
“And even when the returns are immediate, as in the T&TEC research to improve grid quality, bureaucracy freezes the process. In fact, says Hosein, many local companies tend to prefer to get foreign consultants. If TTLAB can continue to grow, this could change, as “we would like to go beyond just publishing, and start filing patents and developing local products to show we can benefit local industry.”
TTLAB’s founder wants to keep it focused on research but admits that some development may be required in the future to help pay the bills. “The preferred model is simple,” he said. “A company admits they need help, provides us with the information (especially data) that we would need, we provide a solution and possibly a prototype of the solution and we move on. Of course during the process we hope to publish but we will never disclose any confidential or company proprietary information.”
The object and means to develop might be science, but the spirit that allows it to exist must be cultural.
• Article courtesy ANSA Caribbean Awards