Centre for Artificial Intelligence and Robotics (CAIR)

Centre for Artificial Intelligence and Robotics (CAIR)

Historical Background
Centre for Artificial Intelligence and Robotics (CAIR) was established in Oct 1986.Its research focus was initially in the areas of Artificial Intelligence (AI), Robotics, and Control systems. In November 2000, R & D groups working in the areas of Command Control Communication and Intelligence (C3I) systems, Communication and Networking, and Communication Secrecy in Electronics and Radar Development Establishment (LRDE) were merged with CAIR. With this, CAIR has become the premier laboratory for R&D in different areas in Information and Communication Technology (ICT) as applicable to Defence.

Today, CAIR boasts of an organisational strength of over 300 personnel including over 150 well qualified scientists.

Centre for Artificial Intelligence and Robotics (CAIR) - Achievements
Over the years, CAIR has developed and delivered to its users, a number of Information Systems, Communication Systems and Security Solutions. While doing this, CAIR has also done high quality R&D work in Intelligent Systems technologies.

Information Systems
Today’s battlefield is characterized by fast tempo and high degree of movement of men and material on a wide front. This demands rapid acquisition, dissemination, collation and fusion of information amongst field commanders involved in tactical planning and battlefield operations. In this scenario, decision making is a time-sensitive and complex operation due to the need for processing multiple streams of high rate data (which could be representing any combination of text, audio and video) originating from various geographically separated sources. This scenario exists even during peacetime operations also, perhaps with less stringent requirements. Command Control Communication (C3I) System consisting of networked computers in the tactical battle area facilitates comprehensive situation awareness by the commanders and timely and optimum decision-making.

CAIR is the developing a series of C3I systems for the different branches of Indian Army.

Communication and Networking
CAIR is engaged in developing technologies needed for futuristic tactical communication network for the Indian Army. The lab is assisting the user in specifying the system architecture and the subsystem specifications for their futuristic “Tactical Communication System” (TCS).
Preliminary R&D activity has been launched in anticipation of the Government of India sanction of a major program for the development of TCS.

Some of the recent Communication Networking products designed and developed by CAIR include

Wireless Message Transfer Unit (WMTU) that enables transmission of IP packets over wired and wireless media using Mil Std 188-220 A protocol. It provides connectivity to mobile clients for accessing network resources

Programmable Communication Interface Unit (PCIU) that provides interconnectivity between a formation HQ LAN and a battlefield-wide WAN. It provides facilities for interfacing to Fibre optic modems, HDSL modems and Synchronous/Asynchronous modems

Information Security
Information security aims at securing confidentiality and integrity of the data stored in computers and communicated over networks.

CAIR has been working on the following areas in this field:
-Secure Communication
-Authentication (including speech biometrics)
-Network access control
-Traffic analysis and intrusion detection systems

CAIR has brought out a number of specific products/ solutions in the above areas and has successfully delivered to the different user communities. These include High speed traffic analysis software, Secure e-mail, Information security gateway, IP security, Network and Host Intrusion Detection Systems, On-line disk encryption (Secure Storage), Smart card based authentication, Multi-layer network Firewall, Personal (host) firewall, etc.

Communication Security
Communication Security products developed by CAIR include devices for link encryption, subscriber end voice encryption, and voice scrambler for HF radio. Recent developments include WAN encryptors, Satellite link encryptors, Wire line encryptors, and Voice/Fax/Data encryption devices for voice band circuits. Proprietary algorithms have been developed for IP security and file/message encryption. All these devices are implemented on trusted platforms developed in-house and built using reconfigurable hardware. All associated device drivers and firmware have been developed in-house.

Intelligent Systems
Intelligent Systems technologies development at CAIR is primarily helps it in keeping abreast of technological advances in the fields of Artificial Intelligence (AI) & Neural Networks (NN), Computer Vision (CV), Signal Processing, Robotics, and Virtual Reality.

AI & NN
CAIR is currently developing a number of Data Mining tools using Artificial Intelligence & Neural Networks and is building software libraries to create a Data miner’s toolbox.

Semantic Web is a new paradigm proposed to make the World Wide Web (WWW) more machine tractable so as to establish a foundation for Agent based technologies. The large amount of information contained in languages used by humans necessitate development of technologies to process the information automatically. Human language processing technologies are being implemented using Commercial off- the- shelf (COTS) computers with huge memories and high processing speeds. CAIR has so far been developing technologies for processing human language inputs namely, information extraction, shallow Natural Language Processing (NLP), Semantic Web tools and techniques, and Semantic Web as a knowledge representation structure for human languages.

Decision Support System Shell architecture named AADARSHA that makes use of Object Oriented concepts has been developed by CAIR. It meets several objectives such as lean and thin shell for easy maintenance, universal interface, easy incorporation of new algorithms into inference engine etc.

Computer Vision
In the area of computer vision, CAIR has developed some prototype products for specific problem solving. Work is ongoing in the area of image registration using 2D, 1D and point features, progressive transmission and compression of images, content based image retrieval and multi-sensor image fusion. Research focus of CAIR in this area has been in registration of multi-sensor images, super-resolution image generation from low resolution images, automatic characterisation of image quality, and characterisation of shapes in grey scale and colour images.

Another important application that is being looked into is the build a mosaic of images taken from various angles. This is for the generation of panoramic and large scale image piecing together many small images seamlessly.

Robotics
CAIR has developed a variety of controllers and manipulators for Gantry, SCARA and other types of robots. These were supplied to Public Sector Units such as HAL and sister DRDO labs. CAIR has gone on to develop a prototype Unmanned Ground Vehicle (UGV) with the aim of attaining autonomous capability. This involved in-house construction of mobile robot platforms, integration of infrared sensors with the vehicle, and the development and integration of path planning software. An useful offshoot of this work was the development of an intelligent wheelchair that would help physically challenged people both in hospitals and homes. One version of the wheelchair could be operated using human voice commands. Another was equipped with a camera system to get information about the surrounding space for its path planning.

Other robots developed by CAIR are for Non-destructive testing, Ammunition loading, and Hot slug manipulation. Both wheeled and legged miniature mobile robots have been developed.

Virtual Reality
Virtual Reality (VR) is an emerging technology with wide applications in product design, operator training, design review etc. The objective of VR is to provide the user with the sensory experience of being in an environment that is not physically present. CAIR has acquired considerable expertise and infrastructure in VR, which could be utilised to develop advanced virtual reality applications.

CAIR has competence in the development of drivers for VR accessories (like flock-of-birds position tracking system, cyber glove, serial ports and stereo drivers for IRIS Performer) expertise in Multigen and Paradigm Vega for modelling and simulating 3D virtual environments, expertise in OpenGL, VRML programming, development of terrain visualisation and flythrough system.

The research on advanced virtual reality applications requires good knowledge in virtual reality modelling and simulation software and programming skill in VR toolkits like OpenGL, IRIS Performer, VRML etc. which have been sufficiently acquired and developed by the scientists of CAIR

A virtual reality theatre consisting of stereo projectors, high-end workstation, 3D glasses and a head mounted display device and other accessories has been installed. Using this facility and other tools such as MultiGen modelling software, a system for simulating terrain flythrough has been developed.

Signal Processing
Signal processing is an important activity at CAIR, with emphasis on speech and image processing.

In the area of speech coding, a Linear Predictive Coder (LPC) based voice coder that encodes 4 kHz band limited telephone quality analogue speech into a 2.4 kbps digital stream has been developed.

In the area of speech recognition, speaker independent speech recognition and language independent speaker identification are active areas of work.