On the labor and delivery floor of Beth Israel Deaconess Medical Center, about 20 so-called ninja nurses use their sixth sense to efficiently assign staff and resources to patients to make sure everyone gets the care and attention they need.
Now, a doctor and an Massachusetts Institute of Technology professor are teaming up to use their medical and robotics expertise to improve how machines work with humans, and make hospitals a bit better in the process.
"What we're aiming to do is learn from people who are outstanding at these resource allocation jobs and potentially teach a machine," said Julie Shah, the professor who leads the Interactive Robotics group at the Computer Science and Artificial Intelligence Laboratory at MIT.
Shah and her husband Neel Shah, an obstetrician at Beth Israel, will spend the next two years learning from these "ninja" nurses — officially called resource nurses — to try to understand how they make certain choices to improve decision making in the hospital and in machines.
The Shahs will develop a simulation test for the resource nurses, with the goal of translating instinctive decisions into specific explanations for certain actions. The project is funded by the Harvard Risk Management Foundation.
"If we can learn what these rules are that the best people are using, we'll be able to train people better," Neel Shah said.
The Shahs said they hope to have a training tool that can help other nurses make better decisions within two years, but eventually hospital floors could have intelligent machines to help hospital staff make decisions.
"Any tools we can give clinicians on the front line and control as best we can are really helpful and (can) make care safer," said Carol Keohane of CRICO, which awarded the grant. "It will help to hone in, and help people identify what resources are needed and take care of this population as well as possible."
There is no intent to take jobs away from hospital staff, Julie Shah said. Instead, the research will be used to help nurses make decisions and train new nurses to have the same "ninja" prowess.
"This is an area where long term it's not practical to have machines doing the work," Julie Shah said. "We still need people doing it, the question is how do we support people doing it."
For Julie Shah, the research will also help with what she calls "re-planning," making decisions and adapting to new scenarios without explicit instructions.
Her research focuses on the decisions that machines — largely robots — make autonomously, without having to be explicitly told to complete a task or alter a plan as well as how machines work with humans.
But, if machines have a better understanding of the decisions that humans make, the machines could re-plan and adapt to changing scenarios better.
Anda sedang membaca artikel tentang
Robots to learn from nurses
Dengan url
http://terdiamtersipu.blogspot.com/2014/10/robots-to-learn-from-nurses_5.html
Anda boleh menyebar luaskannya atau mengcopy paste-nya
Robots to learn from nurses
namun jangan lupa untuk meletakkan link
sebagai sumbernya
0 komentar:
Posting Komentar