New Electronics – Who needs a fitness trainer anyway?
Augmented reality, AI and advances in sensor design mean wearable devices could put athletic trainers on the bench, as Alex Brinkley finds.
Not everyone can afford a personal trainer to stay motivated and on track in the gym or on the sports field. For those who want personalized advice at no cost, the sensor industry takes over. By tackling some of the conventional design challenges, it’s poised to be the next disruptor in wearable technology.
Smartwatches have been monitoring our heart rate, step count, and calories burned since the first FitBit launched in 2009. Smartwatches followed in 2014, and the health market has never looked back.
The design of fitness trackers and smartwatches have been refined over the years, becoming the slim models we know today, but they’re still considered too bulky for many forms of active exercise.
Molex and Avnet commissioned Dimensional Research to survey design engineers for the Future of Diagnostic Wearables: The Future of Medical Monitoring report.
Respondents identified device size, including sensing elements and connectors, as one of the main design challenges. There were also concerns about a simple user interface, power management, signal quality, and thermal management.
A sensor can record changing environments as well as measure pressure and motion. This requires them to be small to minimize the final form factor, but also reliable and stable in operation under different conditions.
The evolution of sensor design has resulted in integrated sensors designed for less bulky end products that can easily be worn by active users.
Take a look at running
French company ActiveLook, a division of OLED micro-display maker MICROOLED, has made great strides in implementing augmented reality in smart glasses. Its head-up display technology is integrated into smart glass frames and lenses to project data onto the lens.
The micro-display module is nearly invisible and does not obstruct the wearer’s view of the horizon. It is based on a monochrome AMOLED screen, with a resolution of 304 x 256 pixels and consumes less than 1 mW, weighs 6 g and has a battery life of over 12 hours.
OLED technology allows connected apps to display data on the lens for the wearer to monitor parameters such as distance, time, speed, pace or heart rate via a smart chip embedded in the frame of the glasses. eyeglasses.
The company has partnered with app providers, such as Openrunner and iKinesis. Openrunner displays real-time routes to navigate as well as running, hiking or biking routes.
The eponymous iKinesis running coach app is connected by a Kapsule, a nine-axis inertial sensor. Athletes attach the Kapsule to the front of the running shoe and the app then analyzes data metrics, such as pronation, kicks, speed, propulsion efficiency and cadence.
The idea is for the runner to focus on the chosen route while receiving real-time data to correct their gait. The sensor communicates via Bluetooth Low Energy 5.
Later, the app also uses data from the Kapsule to reproduce the movement of runners in 3D, to show the center of gravity, the striking attack and the forces involved in specific areas such as the knees, hips and ankles. It also displays the iKinesis index, i.e. distance, speed, duration as well as technique such as stride and running cadence. Algorithms interpret the data to analyze the most stressed areas in order to avoid injuries. An interpretation of running technique can create a prevention routine from a library of exercises categorized into stimulation, stretching, and massage.
The ActiveLook device connects to a smartphone, watch or specific sensors via Bluetooth Low Energy to save power consumption. It has, for example, also teamed up with the Finnish sports watch manufacturer Suunto. Its smartwatch sensors will collect and process data that will be displayed on the connected proximity display. For example, it can send road alerts to be displayed on the glasses, so that the runner adapts his level of effort or changes equipment according to the conditions.
In the swim of things
The obvious design problem for swimmers is to protect the wearable device when submerged in chlorinated water and this tends to be solved with encapsulation techniques.
Bosch-Sensortec, for example, designed the BMP384 pressure sensor in an LGA metal cover with a gel-filled cavity to increase robustness against water, liquids and chemicals. It is compact at 2.0 x 2.0 x 1.0mm and can measure water levels and pressure.
It operates in a range of 300 to 1250 hPa, with a current consumption of 3.4 A and 2.0 A in standby mode.
An accelerometer in a waterproof bracelet can detect the start of swimming and is usually connected to the cloud for analysis. Shot-by-shot feedback, however, requires the collected data to be transmitted and processed remotely, which can be costly in terms of data bills. It also requires the device to be paired with a smartphone or connected to the internet, which can be expensive or inconvenient. Storing and processing data requires both memory and a processor, usually a GPU for speed, which can increase the cost of the end product.
Bosch-Sensortec has developed a swim tracking device using its BHI260AP smart sensor and sensor fusion. The smart sensor integrates a three-axis accelerometer and a three-axis gyroscope with a floating-point microcontroller to provide raw sensor data and perform artificial intelligence functions. Using an application processor reduces the processing demand, thereby conserving system power consumption.
The trackers software uses the date of the motion sensor to determine that the swim session is in progress, without the swimmer having to turn it on or activate it in any way. The software determines which swim stroke is used and records the number of strokes, laps, and pauses between laps. Using the number of laps and strokes with the swimming strokes category allows the wearable device to calculate performance, based on the number of seconds to swim one lap plus the number of strokes. This data allows swimmers to monitor performance and identify areas of weakness. It is at this stage that a human trainer may be needed to perfect the style or technique.
The AI can gauge the swimmer’s height, the length of their limbs, and the speed and power at which their arms and legs move through the water. It can also track any fluctuation in the speed at which limbs move to track fatigue or weak areas.
AI software can be configured based on available compute resources or budget. For example, latency or power consumption can be traded to meet the desired code size.
Another design advantage of using AI software is that it can be configured to follow a specified or adapted standardized swimming style based on swimmers’ ability or performance class; professional or amateur, for example.
Over time, AI could be used to recognize and adapt to swimmer age, size and ability and provide analysis to improve technique and/or performance. For now, coaches can visualize AI-generated 3D data or images, to identify areas for improvement, such as where a running gait might put too much weight on a joint to correct. or to perfect a swimming technique to improve speed, propulsion. or turn.
IDTechEx identified miniature pressure sensors as an additional opportunity for wearable sports devices.
Worn on the wrist or ear, the devices can use these sensors to measure push-ups or strength training for gym goers to track how much they lift without relying on a smartphone.