UCSC-ICTA finger-print innovation saves Rs.1b

The locally developed Automated Finger-print Identification System (AFIS) which was introduced recently could save the Government about Rs. 1,000 million, said ICT Agency of Sri Lanka (ICTA) Chairman Prof. P. W. Epasinghe.

The AIFS was launched by Telecommunication and Information Technology Minister Ranjith Siyambalapitiya and Defence Secretary Gotabaya Rajapaksa at the National Deyata Kirula Development Exhibition 2014, in Kuliyapitiya.

The system which replaces the 106-year-old manual system is a product of ICTA and the University of Colombo School of Computing (UCSC).

Through this system the UCSC and ICTA have facilitated 10-print identification (Convict Identification) quickly and at low cost.

Before the introduction of this innovation, the 10-print finger print police report took time, ranging from one to 14 days.

With the introduction of the ICT leveraged innovation the time taken for a 10-print finger print identification is less than three minutes. With computerisation of the Criminal Record Division (CRD), the process of identification and preparation of report takes less than 30 minutes.

Sri Lanka has been invited to present this system at the National Institute of Standards and Technology (NIST) in Paris in June, said Prof Epasinghe.

"Several new algorithms needed for Sri Lankan conditions were developed for this finger print identification system," said UCSC, Software Unit Head, Harsha Wijayawardhana.

"If we purchased the system from abroad it would have cost the Government a lot. But we adopted a system to suit Sri Lanka. This system also helps maintain confidentiality. To speed up identification, several new algorithms had been used in the system based on parallel computing," he said.

"It is now possible to bring reduce the time taken for recognition by adding a server or servers to the present system. As a result of the use of a server farm or collection of computer servers, the cost is reduced drastically and identification of 10-print is made within three minutes with 99.9999 per cent accuracy," Wijayawardhana said. Within 12 days of the launch of the system, 10-print, finger prints of 4,035 persons were identified and verified.

Of this 899 were identified as being those of known criminals. This is 22 percent of the people identified. Up to 397,000 finger prints collected from crime scenes have been included in the AFIS.