The Robert Bosch Centre draws upon the existing deep strengths within the Indian Institute of Science, Bangalore to drive multidisciplinary research programs in selected application areas involving complex interacting systems, while simultaneously developing key foundational strengths in important core areas within the centre. These research programs will address significant challenges facing our country in particular and the society in general and will look to apply the advances in sensing, computing and analytics to understand and solve problems. The thrust areas will be identified based on the importance, relevance, context and capability and will evolve over time. Currently, we believe programs in smart cities, health systems and autonomous systems are extremely important to the country and also offer enormous opportunities for research, innovation and societal impact.Smart Cities
Smart cities have been identified as a high priority, national mission by the Government of India. India is undergoing rapid urbanization and the only way city infrastructures can keep up is to re-think and apply technological advances to enable rich and satisfying environments for the city denizens. Cities are perfect examples of complex interacting systems, with multiple subsystems working in semi-independant yet collaborative manner.
Take, for example, the mobility system in a city: This enables the movement of goods and people and is one of the most important requirements of any economy. Interactions between traffic systems, road systems, public and private transport carriers and the mobility demand of people and economic entities, all combine in complex ways to determine the performance of this system. Today there is poor understanding of how these systems interact with each other and each is dealt with in Silos. It is currently difficult to answer questions like: What happens to the effective travel time if 10,000 new Tata Nanos are introduced into the city? How would making certain roads one way improve mobility? Many of the subsystems have deep connections and interactions – for example water distribution requires energy. Hybrid vehicles and metro transport too draw from the energy grid, Social events put demands on surveillance and transport systems and so on. Ultimately, human behaviour and choices impacts almost all of these systems.
Our current approach to cities is largely reactive and we are unable to keep pace with their staggering growth. This can only be mitigated via application of Information and Communication Technologies (ICT) for not only data driven management of these systems but also do new science to develop deep understanding of these systems. The scale of Indian cities, with many having 10-20 million citizens, and the need for being extremely cost-effective, offers exciting opportunities for both fundamental research as well as technology innovation. With data being the key raw material for smartness, we need to develop technologies and policies to harvest, manage and mine this data for public good, while maintaining data integrity, privacy yet enable openness . Since this is an emerging area, there is not enough experience, especially in the Indian context, to decide on choices of various technologies, approaches and policies. Hence there is a crying need for a test bed which will allow experimentation and evaluation of various aspects related to smart cities.
We will specifically look at energy, water and mobility systems and their interplays along with research on systems engineering issues in IoT communications, security and distributed analytics. We will aim to put together a coalition of partners, including a partner city, companies and other universities and the state and central government in this endeavour.
There are two ways in which a cyber-physical approach can impact future healthcare needs. One way is through the creation of low cost, low power sensors to enable applications in remote healthcare and continuous personal health monitoring. Typically, such applications are restricted to the measurement of common physiological and biochemical parameters such as heart rate, temperature, glucose, lactic acid and so on. The challenges in this area mostly comprise of designing under constraints of size, ease of use and power. It is our view that problems of this nature are better addressed in a industrial lab setting where professional engineers and product designers can team up to create effective solutions.
The second way in which a cyber-physical approach can impact future health care is by addressing outstanding challenges involving large number of – often non-linear – interactions, interfering with sensing in the low-signal limit or complicating prediction of outcomes of drug treatment and so on. It is the latter path which is ideally suited as the focus area for an academic centre. A few examples of research problems under this category will be robust sensing of low level biochemical indicators from complex environments, understanding mechanisms of drug resistance and so on. We will identify expedition projects which address the issue of complexity in sensing and management of human health.
Systems which are capable of interacting with its environment are often called autonomous systems. It is now globally recognized that these systems will have enormous impact in a wide variety of disciplines which include agriculture, surveillance, healthcare and others. Realizing its importance the Robert Bosch Centre develops the area of autonomous systems. Though autonomous systems can be interpreted more broadly, we will confine ourselves to physical systems which can be controlled remotely. Over the last several decades significant progress has been made in building such systems. However, adapting them to specific tasks such as agriculture still remains a significant open problem.
In this centre we will take systems which are readily available in the market such as drones and study the control and command aspect. Going forward we will study the interaction between several such drones. This will have enormous applications, especially in surveillance and monitoring.
In the medium term we will develop autonomous systems which have cognitive abilities to do imprecise reasoning: Imagine a system which could do question answering.
The centre will draw on the expertise of faculty in the IISc in Control, Machine Learning and Mechanical Systems engineering residing in Departments of Aeronautical Engineering, Mechanical Engineering, Electrical Engineering and Computer Science and Automation.
In addition to these thrust programs, we will also have various other exploratory research threads that involve either applications or foundational aspects of cyber-physical systems, which then could grow into large programs.
The key hallmarks of these complex systems are:
- Complexity: This arises either due to rich behaviour of individual components (for example humans, drones etc.), or the rich interactions between components (human to human or human to machine, wind to turbine etc.), or the scale in terms of the number of components (a city scale traffic system).
- Interactions: Difficulties and interesting behaviours emerge due to the interactions of components. For example in a wind farm, while the wind causes a turbine to move, the moving rotors in turn change the wind pattern which impacts downstream turbines.
- System of systems: Many of these systems can be decomposed into components which themselves will be complex systems. Examples of weather system, ground water system, electric grid and the plant system combine to determine the fate of the crop system.
- Uncertainty and unknowns: The scale and nature of components implies that it is not possible to know everything about the components, nor will it be possible to measure them with full fidelity. A classic source of uncertainty comes from the presence of a human agent, an important component of many of the above systems.
- Cyber-physical nature: With introduction of technologies for sensing and actuation, we convert many of these systems into hybrid systems – part physical and part digital. Thus co-engineering of these becomes an important aspect to understand.
While many of these topics are studied in silos in various departments within the Indian Institute of Science and in other institutions around the world, there seems to be a need to study such systems in a holistic fashion and hence motivates this centre. This approach can explicitly focus on the systems and interactions aspects through theoretical and experimental investigations. Recently many new related centres have come up around the world and we aim to network with these.
The institute will aim to build up core strengths in the following core areas.
- Modeling, simulation and analysis of such complex interacting systems, with focus on data driven models, real time simulation to enable model based closed loop control, agent based models, stochastic models and simulation of large cyber-social systems, and large scale models (for example cities).
- Optimization, control and policy, with a focus on human in the loop systems, model based control, stochastic control, network inference and control, and mechanism and policy based control.
- Large scale sensing and actuation systems with focus on multi-dimensional sensing, energy efficiency, sparse sampling/actuation, crowd sensing/actuation, and feature/information sensing.
- System architecture with focus on large scale distributed systems engineering with emphasis on scalability, energy efficiency, real-time, correctness, security, and data privacy.
Read more about our Research Projects and our Publications.