SHARP logo Hope College, Holland, Michigan
All About SHARP applicants


program directors



All About Sharp
Project details

Sensors and Machine Learning Algorithms to Detect Posture

Project Full Title:

Wearable Sensors and Machine Learning Algorithms to Detect and Classify Posture during Manual Patient-Handling Tasks

Project Mentor(s):

Odle,Brooke; Olagbemi,Omofolakunmi

Project Mentor(s) EMail:;

Project Start Date:


Project End Date:


Project Description:

HOPE STUDENTS ONLY, PENDING FUNDING. This interdisciplinary project will incorporate the disciplines of Nursing, Engineering, and Computer Science. However the project is specifically housed within the Hope College Department of Engineering and Computer Science. There is a prevalence of musculoskeletal injury risk associated with caregivers (nurses, nursing aides, physical therapists, etc) who perform manual patient-handling tasks. These tasks entail lifting, pulling, twisting, squatting, and pivoting. Sometimes they are performed in tight spaces and in awkward postures that place high forces on parts of the body that are especially prone to injury, typically, the lower back, neck and shoulders. Of these categories, injury to the lower back accounts for the highest number of cases of musculoskeletal injuries sustained by nurses and caregivers [2]. In this project we seek to explore the influence of posture on the performance of manual patient-handling tasks, and to determine related metrics that can be used to categorize posture. This will be achieved by implementing experimental biomechanical techniques and machine learning algorithms. The specific aims of this project are summarized as follows: 1. To determine both optimal sensor placement on body parts and an appropriate set up to capture sensor data while participants perform pre-determined tasks. 2. To develop and validate a machine learning algorithm to classify posture (good, poor, neutral) based on sensor signals captured during performance of tasks.

External Link:




© 2007-2010 Hope College, Holland, Michigan, USA. All rights reserved.