STEM II — KinetiStep

Assistive Technology Project — Adaptive Knee Brace for Stair Navigation
Naaisha Agarwal · Rddhima Bora · Rishi Gandhi · Neil Gupta  |  Spring 2026

Problem Statement

As people age, muscles naturally weaken and joints degrade. Approximately 13% of women and 10% of men over the age of 60 develop knee osteoarthritis (OA), and that risk rises to 40% for individuals over 70. Osteoarthritis causes the cartilage in the knee to gradually wear down, leading to bone-on-bone contact and significant chronic pain. While many individuals with OA can manage discomfort during normal walking, stair navigation presents a much greater challenge: stair climbing requires greater knee flexion and produces substantially higher patellofemoral joint forces compared to level walking. Studies show that individuals with knee OA are 50% slower climbing stairs, and they also have 30–38% weaker knee muscles and 17% weaker hip abductor muscles compared to healthy adults, directly increasing pain during stair use.

Given how frequently stairs appear in homes, offices, and public spaces, individuals with knee OA need an affordable, accessible device that helps them navigate stairs independently, with reduced pain and increased stability, without requiring significant arm strength or a fully functional second leg.

40% OA risk for adults over 70 50% slower stair climbing with OA 30–38% weaker knee muscles

Design Approach

Market Research

Existing devices fall into three categories: unloading knee braces, stair-assist canes, and step-resting platforms, each with significant limitations. The Unloading OA Knee Brace relies on manual strap adjustment, is not designed for stair movement, and does not adapt to changes in gait. The Stair Step Assisting Device requires one pain-free leg and strong arms, excluding most elderly users. The Stair Assist Cane improves balance but provides no direct knee support. None of the existing solutions combine adaptive knee support, stair-specific design, independence from arm strength, and affordability under $100.

Preliminary Design Concepts

Concept 1: Pneumatic Brace

Air pockets on the front and rear of the knee inflate alternately during stair ascent and descent, providing assistive pressure with no rigid parts. Fully flexible and breathable. Main concern: durability and pressure consistency over extended use.

Concept 2: Linear Exoskeleton

A compression coil spring stores energy during knee flexion and releases it during extension, reducing the effort required to step upward. Estimated cost $72–82. Concern: added leg weight and reduced effectiveness on stair descent.

Concept 3: Torsion Spring Brace

Two metal plates connected by a torsion spring sit above and below the knee, providing torque opposite the direction of knee bend to ease stair climbing. Portable and low-cost. Concern: rigidity may reduce comfort and mobility.

Final Design: Adaptive Angle Brace

After evaluating all three concepts against our engineering requirements matrix, the team developed an Adaptive Angle Brace: a motorized knee brace that uses real-time angle sensing to provide dynamic, personalized support during stair navigation. The brace consists of two curved plates (one under the thigh, one under the shin) connected by a torque motor. An IMU sensor continuously measures the knee angle. An ESP32 microcontroller processes this data and commands the motor, which adjusts its assistive force based on where the user is in their stair-climbing motion. The entire system attaches to the leg with elastic straps and velcro for easy independent donning and doffing. Electronic parts included an IMU, ESP32, motor driver, and 3.7V battery, all of which are housed in a compact 3D-printed enclosure on the thigh plate.

IMU (MPU-6050) ESP32 Torque Motor LiPo Battery ArduinoIDE (C++) 3D Printed PLA/PETG Neoprene + Velcro

Build Process

1. Sketch & Finalize Dimensions
The team produced a detailed hand sketch of the final brace, labeling all components and dimensions to understand how every part would fit together before committing to CAD.

2. CAD Prototyping
Each component was modeled individually. Curved thigh and shin plates, motor housing, motor arm block, and electronics enclosure were all created in CAD. Shared hole positions were projected across parts to ensure precise screw alignment. All files are on the team's GitHub.

3. 3D Printing
All CAD parts were printed in PLA/PETG. Fit and clearances were verified before electronics assembly.

4. Electronics Setup
The IMU was wired to the ESP32 via breadboard and programmed in ArduinoIDE to compute the live knee angle. Once confirmed, the motor, motor driver, and 11V battery were added. Final code translates knee angle readings into motor commands.

5. Final Assembly
Electronics were seated in their enclosures, all components were screwed together, and the brace was fitted with elastic straps and velcro. The completed prototype was then tested on a client who navigates stairs regularly.

Testing & Results

Pain Reduction

Average pain ratings: 8.4/10 without brace vs. 4.6/10 with brace — a 45% reduction. Paired-samples t-test: p < 0.001. Level 1 criterion (≥10% reduction) met.

Knee Collapse

Average knee collapse angle: 5.2° without vs. 3.2° with brace — a 38% reduction. p = 0.011. Level 1 criterion (≥10% reduction) met.

Knee Mobility

Average mobility: 95° with brace vs. 105° without — a 10° reduction. Target was 120°. Level 1 criterion not fully met; future iterations will use softer materials to improve range.

Comfortability

Average ratings: 9.4/10 at 1 minute and 7.5/10 at 3 minutes across 5 participants. Level 1 comfort criterion met.

Prototype

The final prototype consists of two 3D-printed curved plates strapped to the thigh and shin with velcro, a torque motor bridging the two plates, and a compact electronics housing containing the IMU, ESP32, motor driver, and LiPo battery. The adaptive code runs continuously on the ESP32, reading the knee angle from the IMU and actuating the motor to provide assistive force at the right point in each step.

Sketch, CAD files, and full source code are available on the team's GitHub: github.com/naaishaagarwal-wpi/KinetiStep

Project Poster

If the PDF does not load, click here to open it.