Effective body monitoring systems are indispensable in both laboratory and real-world settings. They must possess the capability to accurately monitor subtle nuances such as aspects of gait, as well as more conspicuous movements like postural sway and specific gestures. Moreover, these systems need to cater to diverse populations, spanning from children to adults, elderly individuals, and those with disabilities. Additionally, they should be adaptable to tasks ranging from short durations to extended periods lasting several hours. Developing systems that meet these broad criteria is inherently challenging.
Consequently, a multitude of systems have been devised, each tailored to address distinct research agendas. These systems can be broadly categorized into four main types: 

• Contact sensors
• Non-contact sensors
• Mixed systems
• Manual approaches. 

Each of these classifications serves different purposes and is tailored to specific requirements. It’s important to note that these systems can also be classified in various other ways, such as by the type of energy utilized or whether the sensor is physically in contact with the individual. Our classification serves as a practical framework for discussion purposes.

Contact sensors
Contact sensors encompass two primary types. Firstly, there are body-worn accelerometers, magnetometers, and gyroscopes. For instance, Feese utilized a set of six inertial measurement units positioned on the arms, back, and head to capture posture mirroring during a social interaction task. Data collected during this task was leveraged to develop an automated classifier for three basic arm positions (left arm up, right arm up, both arms up).
Due to the tendency of inertial sensors to drift out of calibration, they are often affixed very closely to the body within skin-tight garments. Sensor-integrated garments can be highly effective, although their efficiency is contingent upon the fit of the garments to the user. This necessitates the availability of customized garments when monitoring multiple individuals. Unlike some vision-based approaches, they are not affected by changes in illumination and can target specific limbs or areas of the body.
However, there are potential drawbacks associated with contact sensors, including their cost, social and physical discomfort when worn, and high power consumption. Nevertheless, efforts have been made to mitigate these drawbacks by reducing the cost and power consumption of sensor-integrated garments.
The second type of contact sensors encompasses garments crafted from elongation-sensitive yarns, conductive elastomers, or other bend sensors. For instance, Tognetti integrated conductive elastomer sensors into a fabric glove and incorporated them into the arm and shoulder areas of a shirt. They then developed software capable of recording predefined postures, recognizing these recorded postures, and playing them back using 3D animations.
These types of garments typically fit snugly against the skin because they rely on the body’s movements to deform or strain the sensors embedded within the fabric. However, this snug fit may impact both social and physical comfort. To address this, efforts have been made to enhance physical comfort by incorporating plastic optical fiber sensors.

Non-contact sensors
Non-contact systems primarily rely on camera-based technology. These systems capture video footage of individuals for subsequent movement analysis. An example of such technology is the Microsoft Kinect, which has been utilized for near real-time full-body motion capture and gesture recognition. The Kinect is equipped with a color camera, an infrared projector and camera that employs structured light to calculate depth, and a microphone array. Particularly, the depth camera has demonstrated significant utility.
For instance, Biswas and Basu utilized the depth camera to train a classifier to recognize a set of eight human gestures. Similarly, Burba utilized it to detect subtle nonverbal behaviors such as respiratory rate and fidgeting behavior. These examples illustrate the Kinect’s effectiveness in body movement tracking. However, despite its relatively low price point as a mass-market device, the Kinect shares many features with specialized hardware found in commercial systems.
There are several limitations to consider. Firstly, the Kinect cannot analyze pre-existing video footage; data must be collected using the sensor. Secondly, specific system requirements, including CPU, RAM, and video card specifications, must be met to use the sensor effectively. Additionally, while other sensors may be smaller, more discrete, and less distracting than the Kinect, it’s important to note that it may not be as user-friendly. Although free software for skeleton tracking exists, it is not a “plug and play” solution, which may pose challenges for some users.

Mixed systems
Mixed systems utilize a combination of contact and non-contact sensors, such as high-speed kinematic systems and pressure or force sensors. Commercially available kinematic systems like Vicon Motion Capture and Optotrak Certus employ high-speed cameras (non-contact) to track markers placed on objects or individuals (contact). Operating in three dimensions, these systems offer high resolution and accuracy, making them a preferred choice for those with sufficient budget. However, they require meticulous setup and calibration to unleash their full potential.
Alternatively, some mixed systems utilize pressure or force sensors positioned on the seat and back of a chair. Commercial options include the Body Pressure Measurement System (BPMS) from Tekscan and standard force plates. To address cost concerns, lower-cost alternatives have been developed. For instance, Olney and D’Mello created a system using two Wii Fit game controller boards, providing 8 pressure data streams compared to the Tekscan BPMS’s 38×41 sensing array. Despite being cost-effective, challenges such as difficult or impossible calibration and power consumption issues may arise, particularly with the use of Bluetooth communication.
Similarly, Arroyo utilized low-cost/low-resolution pressure-sensitive pads for a chair but encountered practical limitations, restricting sensor usage to about half of their data collection period. Even in cases where resolution remains high, such as with Kamiya’s pressure-sensing array employing 64 Tekscan Flexiforce sensors, cost is a concern, and challenges persist in constructing a reliable, effective system.

Manual approaches
Manual approaches to body movement analysis typically involve trained human coders observing individuals’ movements and making judgments about their actions. This can occur in real-time, but more commonly, videos of movements are recorded for subsequent analysis. For instance, Friesen, Ekman, and Wallbot trained judges to classify hand movements from videotaped conversations into three categories: speech illustrators, body manipulators, or actions conveying symbolic information.
Various coding schemes are utilized for this purpose, such as the Body Action and Posture Coding System (BAP) for classifying body movements and postures, and the Davis nonverbal state and nonverbal communication scales for coding nonverbal aspects of communication and movement behavior, among others. While these systems typically demonstrate high inter-coder reliability, the coding process can be labor-intensive in terms of human capital. In some cases, videos are coded frame-by-frame to ensure accuracy.
While the various approaches discussed offer viable solutions for monitoring body posture and movement, each comes with its own set of advantages and disadvantages. Researchers must carefully consider these factors when selecting a system that aligns with the requirements of their specific research questions.
Contact sensors and mixed systems offer precise measurements at high temporal resolutions, but they can be costly, intrusive, and may necessitate sophisticated hardware and software. Non-contact vision-based systems address some of these challenges, but their effectiveness is often contingent upon favorable lighting conditions and optimal camera positioning.
Manual coding methods are technologically inexpensive and non-intrusive, but they can be labor-intensive and subjective. Even in scenarios where a universally applicable wearable sensor garment or a low-cost fabricated pressure sensor is available, additional considerations such as limited portability, computational power requirements, lower precision, scalability limitations, the need for specialized training, and the risk of sensor damage must be taken into account. These factors must be carefully balanced against the benefits they offer in addressing specific research questions.

How these technologies are shaping various industries
In recent years, advancements in body motion tracking technologies have revolutionized various industries, from healthcare to gaming and beyond. These innovative technologies have paved the way for transformative applications, offering unprecedented insights and capabilities. Let’s delve into how these advancements are shaping industries worldwide.

Healthcare
Body motion tracking technologies have found extensive applications in healthcare, particularly in rehabilitation and physical therapy. These systems enable precise monitoring of patients’ movements, allowing therapists to track progress, customize treatment plans, and provide real-time feedback. Additionally, wearable motion sensors are being utilized for remote patient monitoring, empowering individuals to manage chronic conditions and recover from injuries more effectively.

Sports and Fitness
In the realm of sports and fitness, body motion tracking technologies are revolutionizing training methodologies and performance analysis. Athletes can now leverage wearable sensors and motion capture systems to optimize their techniques, prevent injuries, and enhance overall performance. Coaches and trainers utilize these systems to analyze biomechanics, identify areas for improvement, and tailor training programs to individual athletes’ needs.

Entertainment and Gaming
The entertainment and gaming industries have seen a significant impact from body motion tracking technologies, most notably with the introduction of motion-controlled gaming consoles. These systems allow users to interact with games using natural body movements, providing a more immersive and engaging experience. Moreover, virtual reality (VR) and augmented reality (AR) applications are incorporating body motion tracking to create lifelike environments and enhance user interactions.

Education and Training
In the field of education and training, body motion tracking technologies are transforming learning experiences and skill development. Virtual training simulations equipped with motion capture capabilities enable hands-on practice in various fields, such as surgery, aviation, and industrial operations. Additionally, interactive educational tools utilize body motion tracking to engage students and facilitate experiential learning.

Industrial Applications
Industries such as manufacturing and construction are benefiting from the integration of body motion tracking technologies into workplace safety and efficiency initiatives. Wearable sensors and motion tracking systems enable real-time monitoring of workers’ movements, helping to prevent accidents and optimize ergonomic conditions. Furthermore, these technologies are utilized in ergonomics assessments and workflow optimization to enhance productivity and reduce occupational hazards.

In conclusion, body motion tracking technologies are driving innovation across diverse industries, offering unparalleled opportunities for improvement and growth. As these technologies continue to evolve, we can expect further advancements and applications that will redefine the way we work, play, and interact with the world around us. Embracing these advancements will be key for organizations seeking to stay ahead in today’s rapidly evolving landscape.