URGEONS once had the ultimate hands-on job: spreading the rib cage to gain access to the heart, feeling for lumps in a deflated lung, sewing up blood vessels.

But for some operations, like closed-chest coronary bypass surgery, doctors now station themselves not by the patient but at a nearby computer console. With their eyes fixed on a monitor, they use joystick-like controls to guide robotic scalpels, scissors and high-resolution cameras that have been inserted in the patient's body through keyhole-size incisions.

However precise and tremor-free these robotic tools are, though, they lack one attribute that surgeons prize: a delicate sense of touch.

Robert D. Howe, a professor of engineering at Harvard University, wants to change that by creating roboticized instruments that will, among other abilities, palpate internal tissues.

"I think that sensory feedback is the key to the next generation of robotic-assisted surgery," he said. "Current robots have no sense of touch built into them — everything is done purely on visual information."

Dr. Howe spoke last month at a symposium on the future of minimal-access surgery at Columbia University. It was sponsored by New York-Presbyterian Hospital and the Fu Foundation School of Engineering and Applied Science at Columbia.

Minimally invasive surgery — the medical field in which doctors no longer directly touch or see the tissues on which they operate — offers advantages. The tiny incisions limit damage and help speed recovery, and the roboticized instruments that do the surgeons' bidding can scale down the doctors' motions, doing precise work in tight spaces without any hand tremor.

"But the robot can't do everything," said Dr. Michael Argenziano, who spoke at the conference. Dr. Argenziano, an assistant professor of surgery at Columbia University's College of Physicians and Surgeons, is currently testing a system for closed-chest coronary bypasses.

"The robot is an advanced tool for getting into the patient through the smallest incisions possible with the least trauma," he said. "We're going to need a host of ancillary techniques for working in a closed chest."

Dr. Howe hopes to add fine tactile sensitivity to future robots' arsenals. "Operating by eye requires attention and it slows things down," he said. Tactile feedback would make the robots easier to use and speed surgical procedures.

To that end, Dr. Howe's group has developed a device for remote palpation that can sense when one area of soft tissue is harder or rounder than the surrounding tissue. "For example, the surgeon may know that the lump exists because of a CAT scan," he said, but in surgery the lung deflates, making the lump harder to find.

The hidden lump is not a problem in open procedures, in which surgeons spread the ribs and use their fingers to find it. But surgeons who work with robotic probes and cameras inserted through small incisions may have difficulties.

"I've seen surgeons spend 15 minutes finding a lump using a probe passed through a one-centimeter incision," Dr. Howe said. "It's like trying to find a pea inside a bowl of jello using chopsticks."

Dr. Howe's solution is an engineered stand-in for the surgeon's palpating fingers. At one end, deep in the body, is the electronic "fingertip," actually a square matrix of 64 pressure sensors. This fingertip sweeps the interior surface. When it hits a lump, the pressure rises in some or all of the sensors, for the lump does not compress the way the surrounding tissue does.

The other end of the fingertip, which extends outside the body, is a tactile display of tiny motorized pins that the doctor's real finger rests against. Each pin on the display corresponds to one of the 64 sensors at the other end. A processor translates the pressure detected at each sensor into the rise and fall of each pin, so the doctor can feel the contours of the object inside the body.

The device, now only a prototype, is not meant to be a stand-alone instrument. "The key will be to build these distributed tactile devices into surgical tools like graspers and retractors," Dr. Howe said, so that they can be used to locate hidden matter.

Dr. Michael Mack, a cardiac surgeon and chairman of the Cardiopulmonary Research Science and Technology Institute in Dallas who spoke at the conference, said he would welcome tools that could palpate for minimal-access surgery. "Adding the ability to feel is a good thing," he said.

But Dr. Mack, who took part in the initial Food and Drug Administration trials of one roboticized system, said he was skeptical about the current generation of roboticized systems for coronary bypass operations. "Eventually computerization will make it better," he said. But he said that for now, surgical robots offer few if any advantages.

Mastering such instruments involves a significant learning curve, he said. "The second and third generations should address issues that make the systems more friendly."

Dr. Howe said that a sense of touch would be one improvement to robotic systems. "Surgeons will not have to pay so much attention to visual cues as they proceed," he said. "This will leave them extra mental attention for other problems."

Dr. Richard M. Satava, a surgeon and professor at Yale University who discussed the use of smart instruments at the meeting, predicted other changes in operating rooms.

He suggested that many scrub nurses, for example, would eventually be replaced by robotic systems. "No tool will need to be handed to a robot by a scrub nurse," he said. "There's not a lot of thinking in handing an instrument to a robot — that job doesn't have to be done by a person."

Dr. Satava predicts that not only the scrub nurse but also most of the surgical team will be replaced by robots. "We fly jet airplanes by wire," he said. "One day we'll do the equivalent to that in the surgery."