Cloud computing: an e-Infrastructure for collaborative research

Research institutions and organisations are looking beyond their own walls for innovation, opting to create research collaborations to cope with the challenges of a rapidly changing world. Such research collaborations are joint ventures involving two or more organisations with the common objective of producing new scientific knowledge by collating their individual strengths.
Collaborative research involves sharing ideas and concepts through personal meetings and electronic communication; fabricating and materialising those ideas; implementing the ideas in parallel; sharing the results; testing the products; and finally, releasing those products. This scenario becomes more complicated when the collaborating parties are distributed, occupying different geographies and time zones. Thus, a research partnership requires good communication, fast data sharing and high processing power for complex operations. Cloud computing is emerging as a suitable platform to meet these demands.

What is cloud computing?
Cloud computing provides compute resources – such as scalable virtualized servers, data centres, and economical bandwidth – over the Internet. In addition, clouds can provide complex and traditionally expensive database services, such as automatic data indexing data and querying. Major cloud service providers – such as Amazon, IBM, Microsoft and Google – provide these resources using the Infrastructure as a Service (IaaS) model, one of three types of cloud service (Platform as a Service (PaaS) and Software as a Service (SaaS) being the other two). Cloud resources are available on a pay-per-use basis: the consumer pays for the amount of resources they use and the duration of their usage.
Cloud computing assuages the initial investment required to set up hardware, software and networking equipment, and also reduces maintenance costs. It allows for dynamic scaling, such that computing capacity can be increased or decreased as required. Clouds also avoid common problems such as poor network connection or hardware failure since they distribute work across multiple computers. Further, cloud computing can reduce the cost of running applications with lower access rates, since cloud servers can be automatically started on request, or stopped after certain duration of idleness, hence reducing the duration of resource usage.

Cloud usage in collaborative research
Cloud computing can support collaborations between institutes and businesses. It can provide a platform for communicating and exchanging ideas, thus helping organisations of all sizes to work together and beyond their individual boundaries. For example, according to a Reuters’ article, Sogeti Groups, a leading provider of IT services, used ‘IBM idea factory for cloud computing’ to run the first collaborative brainstorming session of its kind.
In collaborative research, clouds also permit collaborative development and/or testing, allowing each of the collaborators to access the product centrally, without keeping multiple copies at each location. For example, TCS Innovation Labs is working with a research partner to create an advanced visualization portal for social video documentaries. While TCS has been responsible for technology, its partner has been responsible for content creation. A convenient approach has been to provide a set of powerful content creation and visualization tools on a shared server, accessible to both. Amazon Web Services (AWS) was selected as the platform for this, providing cheap storage costs and dynamic scalability of processing power and storage; this allowed the incremental addition of content and sophisticated video processing tools, without requiring upfront capital investment from either partner. Another advantage of such shared environments is the speed of content ingestion and technology feedback. Clouds also reduce the costs of travelling to meet partners, and facilitate remote and mobile operations.

What does the future hold?
Cloud computing is evolving and is predicted to have a great future. We can expect partnerships between different companies to add a new dimension to cloud services. For instance, Yahoo and TATA Computational Research Laboratories (CRL) recently announced an agreement to jointly support cloud computing research in India: CRL’s supercomputer ‘EKA’ (the world’s fourth fastest supercomputer) will run Yahoo’s open-source distributed computing software to support data intensive computations.
However questions still remain regarding cloud’s data security. Currently, there is no provision for inter-cloud operability of an application; such a feature would allow users of a single application to choose between various services from different providers. With the evolution of cloud we can expect the present services to grow, with many new and attractive features enhancing cloud computing as an e-infrastructure for research and business solutions.