Breadcrumb navigation

Using NEC’s Gait Analysis Technology to Promote a Healthy Walking Posture: A-RROWG Walking Sensing Insoles

Featured Technologies

February 21, 2020

A-RROWG Walking Sensing Insole is jointly planned and developed by NEC, FiNC Technologies Inc. (Note 1), and Makuake, Inc. (Note 2) and reached its target in just a few hours when supporters were recruited through the Makuake service, on which people make purchases to support new things and experiences. The service has already garnered a total amount of supportive purchases exceeding eight times the target (as of January 2020). What kind of product is A-RROWG? We interviewed the researchers.

To keep walking healthily and beautifully forever

Kenichiro Fukushi
Kenichiro Fukushi
Assistant Manager
Data Science Research Laboratories

― What kind of product is A-RROWG Walking Sensing Insole?

Fukushi: It’s an insole that can analyze the gait of the user, which you can say is the quality of walking. It comprehensively analyzes the walking speed, stride, ground contact angle, ground-off angle, height of raised leg, and extorsion distance to determine whether the user’s gait is healthy and ideal.
As a product, it consists of a pair of dedicated insoles and a smartphone app. Data collected from the acceleration and angular velocity sensors embedded in the insoles is uploaded to cloud via the smartphone app and used for high-precision analysis of the gait. The analysis result derived chiefly from NEC’s gait analysis technology is counted up as points as “gait analysis score” every day and provided as feedback to the smartphone app along with advice for better gait and optimal training videos. It is a product that improves your walking posture and makes for a healthier life while people are just doing everyday activities. We have FiNC Technologies Inc.’s support for gait detection and training advice.

― What were the purpose and reasons for working on the improvement of walking posture?

Fukushi: Actually, before getting involved in this product development, I had opportunities to visit various departments at hospitals. The most memorable of those experiences was meeting all the patients in the orthopedics departments who were unable to walk. The hospital staff said that once old people become unable to walk, they lose the motivation to walk and become confined to bed, which may lead to dementia. This experience made me realize just how important daily walking is to us, and I started looking more into it. Like how the term called “frail prevention” (Note 3) is attracting attention lately, it started out as an idea to develop healthcare solution that can maintain a healthy state before one starts having trouble walking.

Realizing high-precision analysis with a super-compact device

Hiroshi Kajitani
Hiroshi Kajitani
Data Science Research Laboratories

― Why did you make it an insole-shaped device?

Kajitani: For sure, talking about wearable device for the feet, there are ways like directly implanting the devices into the midsoles of shoes or attaching them to the outside of the shoes. One of the big reasons why we chose insole is to make it versatile, being usable with various shoes. Our biggest reason was that we had the aim of making this device usable while people’s natural walking senses are unaffected by embedding it in the thickest part of the insole, specifically, the arch of the foot.

Huang: Just imagine. If this device was worn in a place where people could easily see it, they would be tense from constantly being aware of the measurement. Wouldn’t that make their walking posture unnatural instead? The concept of this product was that the device has meaning only when users can use it on an everyday basis without being aware of it and the device measures the natural walking posture. Therefore, we decided on the insole type.

Kajitani: When we thought about having people use it on a daily basis without stress, we decided that the hassle of charging should be avoided as much as possible. For example, if users had to charge them every one or two weeks, they may do it the first couple of times out of curiosity, but after that they may just leave them inside their shoes upon it becoming bothersome to charge them. Therefore, we aimed for a device that can naturally perform measurements by just leaving it inside a shoe with no need for charging.
The thin, small size requirement also created challenges. Because the battery is small, it can die in short cycles. This device significantly cuts down on energy use and has a level of performance that allows users to keep using it for a year―theoretically speaking―without charging.

Huang Chinhui
Huang Chinhui
Assistant Manager
Data Science Research Laboratories

Huang: To achieve energy savings, we cleverly applied the algorithm that identifies walking and non-walking times. It detects in real-time the characteristics of stable walking from the acceleration and angular velocity data and is designed to start measurement only when the user is walking. It does not falsely sense motion when the user is just sitting and moving his legs, or going up stairs, or on a train. In fact, the device stays in sleep mode any time the user is not walking. This made possible the year-long life of the battery.

Fukushi: In addition to energy-saving performance, this device features NEC’s proprietary high-precision analysis. Generally speaking, acceleration sensor and angular velocity sensor data is considered difficult to handle due to it being prone to error accumulation.

Kajitani: If you simply look at the collected raw walking data, it looks like the user is flying into the sky while walking. That’s how much error can occur.

Fukushi: So we focused on the cyclic motion that occurs during the brief time when a foot comes into contact with the ground, and designed a mechanism that appropriately corrects for noise. Researchers usually perform this process by transferring the data to a research computer. However, this device manages this correction processing using a microcomputer built into the insole. I don’t know any other device that can analyze gait with such precision.

Sensor data and result of gait cycle detectionzoomLarge view
Sensor data and result of gait cycle detection
Comparison of foot trajectory measured by MoCap and A-RROWG
Comparison of foot trajectory measured by MoCap and A-RROWG

Possibility expands with holistic sensing and collaboration with different environment data

― Can you tell us about the future goals and prospects of this technology?

Fukushi: As one big goal, we want to change the prevailing belief that “people can’t walk when they get older.” Our ultimate purpose is to improve users’ gait and extend their healthy life. Nevertheless, we are not hung up on just insoles as a vehicle to that end. We are also eying approaches to build up our original database by collaborating with not only smartphones, but also with wearable devices such as smart watches.

Huang: Yes. For example, if we can comprehensively combine various kinds of data, such as inclination of road surface, information on whether the road is wet with rain, and the user’s health condition, we can see each individual’s lifestyle in a multi-dimensional way and give more accurate, relevant advice.

Kajitani: At the same time, this project is part of our challenge to sense physical motion―research that we’ve been working on for years. I believe that we can draw on the knowledge we gained from the insole and open up possibilities for a wider application by sensing not only feet, but other parts of the body.

Huang: If we can sense the whole body, we can also measure athletes’ movements and visualize them. That may be useful for children’s sports training. Also, it may be applicable to passing on the knowledge of experienced workers at factories. For example, I have heard that painting automobiles and ships require sophisticated knowledge. In this case, we need to sense the painting movements of the hands. If we can visualize the movements of expert workers as data, techniques can be smoothly passed on.

Fukushi: Even when further automation and mechanization awaits in the future, I think there will always be scenes where human experience and intuition are necessary. For such cases, visualizing the accumulated human-based knowledge and passing it on will become increasingly important. Taking this insole as a starting point, we would like to continue working on human body sensing and the research into technologies that can help society.

Relevant Laboratories