IBM Israel’s labs employ 600 people in three separate divisions – the main research lab, the chip and electronic components development lab and the software development lab.The latest accomplishment by IBM lab technicians in Haifa allowed the company’s local management to breathe a sigh of relief. For years the R&D workers, whose labs are attached to Haifa University, have enjoyed high status as “Big Blue’s” largest R&D facilities outside the United States.
Now, with the world headquarters’ decision to increase its R&D in India and China via the recruitment of thousands of workers who will be employed in vast labs, the Israeli labs position could be undermined. The competition within the organization is constantly growing, and any achievement in research or development assumes greater significance.
Thus, when one of the Israeli research groups managed to develop search technology that caught the eyes of top executives at world headquarters, the joy at the Israeli labs was immense. When world IBM decided to develop the technology into a complete product, and turn the Israeli research team into the world team for developing the product, then Israel saw this as an accomplishment that justifies the prestigious status of the Israeli labs.
After the conclusion of the research stage the technology is transferred to the Far East for development and for the manufacture of the product.
IBM Israel’s labs employ 600 people in three separate divisions – the main research lab, the chip and electronic components development lab and the software development lab. The team that developed the search technology is headed by two women, Dr. Yoelle S. Maarek and Dr. Aya Sofer and had until now belonged to the main research lab.
Following the HQ decision to turn the technology into a product and keep its development in Israel, however, Maarek, Sofer and eight of their development people will be spun off from the research division and will form a separate development lab.
The search technology developed by Maarek and Sofer is for optimizing databases. The technology is designed to significantly improve the performance of the BD2 database, which is used mainly by large organizations.
“Of course this is not search technology that will compete with Google or Yahoo,” explains Sofer, “but rather a search engine designed for finding information within organizational systems.
Searching inside an organization is much more complex and complicated than searching the whole Internet – if a search on the Internet misses one result it’s not so terrible, one can always find something similar that is relevant. If, however, you are looking for something within an organization and you miss a critical document that you need for negotiations – it is definitely a problem.
“The search has to be much more precise and comprehensive,” continues Sofer. “The organization is also the only place in which the content creators are the ones who receive the results of the search, such that if the search misses something, the content creators pick up on it immediately. That is the challenge.”
It was this challenge that led to the development of “Guru” – the search engine that is basically the first product of the lab that has come the whole way, from the pure scientific research stage, through the development stage to the promotion of the search technology team to an independent product team.
Along the way, Guru won first place last year in the international search engine evaluation competition, conducted by the American National Institute of Standards and Technology (NIST). “The Israeli research team currently accounts for a little more than 10 percent of IBM’s global research body,” explains lab director Dr. Michael Rodeh.
“We have always been relatively large, considering the size of Israel compared to that of the other countries around the world in which IBM has operations. “The Israeli reality always poses a special and problematic situation, so if we were not extremely good, we would not be here at all. The reason we were given a chance to develop is not due to our size in the Israeli market, but rather the added value that the lab brings to all the company’s products worldwide. Our target markets are always beyond Israel’s borders.
“One must realize that IBM’s massive entry into India and China just now stems not only from the fact that there is cheap labor there – these are markets with tremendous potential for the sales of future products. The company therefore has to adapt specific products to those markets in order to increase sales. We, unlike them, do not have such a giant market.
We know how to operate without a market and to create value via innovative technologies. “On the other hand,” notes Rodeh, “the danger is always there, and we are aware of it. However, since the number of lab workers has increased every year since 1972, even during the recent recession years, we believe that we will manage to create value in the future too.”
One of the recent developments on which the lab worked very hard is the expansion of the European standard for voice recognition by mobile devices. The development in Haifa focused on the addition of functions to the rebuilding of speech and the recognition of the tone and sound of language.
IBM Haifa’s contribution enables computer servers to rebuild the compressed voice that was transmitted to it – in order to reproduce it for the human ear. Until now the possibility of reproducing the original voice and playing it again was not an option under the previous standards.
IBM’s labs worked on this problem with technicians from Motorola, and their partnership recently led to the acceptance of two new standards that will expand the specifications framework approved by the European Standards Institute. In addition, the Haifa labs recently initiated research cooperation with a few international companies that are IBM’s business partners.
The new cooperation model is unlike any that has been undertaken before – whereas cooperative trade activities are an accepted thing, cooperation in research is unheard of. IBM Israel’s R&D labs believe that it will actually be this model that will lead to another breakthrough by Big Blue’s local branch.
(Originally appeared in Ha’aretz)