Unwired. EMTs responding to automobile accidents one day may carry tiny, lightweight devices capable of monitoring each patient and transmitting vital signs to a central computer that continually assesses each person’s condition. Taking into account conditions at nearby hospitals, the computer then will decide where each patient should be taken and will help determine what care needs to be provided on route.

Mark Gaynor, an SMG assistant professor, and his colleagues at Harvard University have developed such a system. Called Hourglass, it consists partly of networks of tiny wireless computers — so small they are called motes, or smart-dust — that monitor patients’ vital signs. Although individual motes have limited sensing and computation capabilities, when networked together and communicating wirelessly, they produce accurate data about their environment, be it the condition of an accident victim, or with different applications, an office building’s air quality, the safety of a bridge, or the soil conditions in the rainforest.

A challenge in developing Hourglass, according to Gaynor, was devising a way to integrate data from the wireless sensors with computers networked via the Internet. The data produced by the medical sensors, for example, need to be integrated with information from medical databases, laboratory results, hospital census data, and data input manually by physicians and nurses.

The researchers expect that Hourglass will not only improve the quality of care in medical emergencies, but also lead to better efficiency and customer service in manufacturing and retail operations by allowing advanced monitoring of supplies, components, and final products as they move through factories, stores, and even into use by the consumer.

This research was published in the July/August 2004 issue of the Institute of Electrical and Electronics Engineers’ IEEE Internet Computing and is available online at: http://people.bu.edu/mgaynor/papers/ieee-grid.pdf.

Pricey spam. A new system devised by Marshall Van Alstyne, an SMG associate professor, promises to help eliminate spam from your inbox — or at least put some spare change in your pocket. His Attention Bond Mechanism (ABM) requires strangers e-mailing you to make a simple promise: that their message isn’t spam. To back it up, those you don’t know contacting you post a small sum, say a nickel, which is held in escrow by a third party, such as an Internet service provider, before sending unsolicited e-mail. It’s a promise that old friends can afford to make, but spammers cannot.

Each subscriber to ABM would create a list of friends, relatives, and coworkers who don’t need to post a bond to e-mail the subscriber, and set the price for strangers. When an e-mail arrives from a sender who posted the bond, you open it, decide whether it’s spam, and if it is, you collect. If it’s not spam, the bond expires and goes back to the sender. The escrow accounts needed to make the system work, says Van Alstyne, can be adapted from those already in use in online auctions such as PayPal.

Other approaches for beating spam have serious weaknesses, he says. Makers of e-mail filters are engaged in a seemingly endless race with spammers who continue to find ways to elude ever more sophisticated filters. And taxing all e-mails, as has been proposed, would put a burden on everyone, not only spammers. Legislation involving do-not-spam lists, ID tags, and labels, says Van Alstyne, would be nearly impossible to enforce outside the country, where many spammers operate.

Van Alstyne developed ABM with former graduate students at the University of Michigan’s School of Information, where he worked before coming to BU. They presented ABM to Microsoft last December and to the office of Senator Ron Wyden (D-Oreg.), coauthor of last year’s Can Spam Act, and the Federal Trade Commission on June 10. The system is described in a paper forthcoming in the Berkeley Economics Journal and is available online at http://www.ssrn.com/abstract=488444.