Today’s workers are inundated with data that is, for the most part, useless. That is not to say it has no absolute use, but rather that it has no use to its specific recipient. Spam notwithstanding, how many of the dozens of messages in an e-mail inbox are of use?
At an RBC e-business conference last month in Toronto, Robert Bauer, chief technology officer with Xerox Global Services, talked about the work his team is doing to reduce the digital clutter.
“No one is getting rid of anything today; there are 150 versions of every document,” he said. “I often talk of that as the rubble of the digital landfills.”
Bauer, who has spent the last 30 years at Xerox’s famed PARC labs in Palo Alto, Calif., said the key to getting use out of all this information is finding a way to structure unstructured documents so they can be machine read. Researchers at PARC are currently working on what Bauer referred to as a smart copier, one which could take 30 disparate papers, images included, and merge them into one summarized document with 85 per cent of the accuracy of a human editor.
The disconnect between man and machine is magnified by the fact that the latter is ruled by databases, the former by documents. The plumbing underneath the ability to freely move from structured data to unstructured is built with XML, he said.
But even if all unstructured data could be searchable, or structured, it is of limited use since people often don’t know what they should be looking for.
Though still years away from truly intelligent documents, those that let users know they are there, progress is being made. Documents are now being structured so they are no longer on or off, accessible or not. “You need to have fine grain permission on individual content,” Bauer said.
An example would be an airplane-parts document where sales would see prices while engineering would see blueprints. The IT industry has to overcome traditional thinking of “trying to make machines work better, not necessarily people’s jobs easier,” Bauer said.
In a subsequent talk, John Parkinson, chief technologist with New York-based Cap Gemini Ernst & Young, focused on some of the hurdles that face technologies developed at research labs such as PARC, before and after they get to market.
Some basic concepts about inventions still exist, he said. “Most innovations are new to the customer, not the world,” Parkinson said. They are not “true innovations but creative use of existing technology.” For instance, the airplane was an innovation, the minivan was not.
But to foster innovation, a company needs to have the right environment and combination of employees, Parkinson said. To do so, there needs to be a bridge between the senior types who think they have seen everything and are constrained by their cynicism, and young employees who are unconstrained by business realities.
Jackie Fern, vice-president and research fellow with Stamford, Conn.-based Gartner Inc., points to the radio frequency identification (RFID), which many companies are talking about placing on all consumer products (disarmed after sales) to track them through the supply chain process, as a hot topic prone to generate vehement opinions.
“There is the ‘big brother is watching’ kind of fear…(but) it is getting a lot of people excited,” she said.
Yet often the resistance and reactions are illogical and unfounded.
Back in the early 1980s Chrysler toyed with the idea of a four-door minivan. Customer focus groups pointed to a failed venture since 80 per cent of current owners (of the three-door variety) said they would not buy a four-door minivan. Trusting their guts, Chrysler management went against the commonly held belief that the customer is always right. Eighty per cent of new minivan orders were for the four door.
The rationale behind this anomaly? “You can’t trust people to make decisions about experiences that they haven’t had,” Parkinson said.