The Algorithm Also Needs Help

The point that I always point out with sensitivity is the fact that the mattress is a personal product. Science has demonstrated through randomized studies that our body structure and the distance between limbs are extremely important in choosing a mattress. These studies have proven that muscle, bone, subcutaneous adipose tissue, body weight and distribution of weight in the body are extremely important when choosing a mattress. It is very important that we choose the suitable mattress for our sleep patterns, sleep habits, sleep problems and sleep time.
Basic factors such as height, weight, body type, and sleeping pattern are the starting point of existing algorithms. Algorithm construct aims to respond in the fastest way with as little information as possible. Therefore, the algorithm should not be expected to perform a correct analysis.
As always, I have to say that the majority of algorithms are used to make perceptions, except for the people who add value to their work and the people who do their job well. For this reason, algorithms are referred to agencies, not to people who are masters of the business.
I do not want to make a feeling that the current applications are inadequate, computer technologies cannot be used in mattress selection. So, what I am trying to say is that there is a lack of parameters and a lack of information on how to set up existing parameters. The application that emerged due to this deficiency is presented as the right option. We need to be determined to provide the truth and empower people to give the right support. If we empower people in this way, it is possible to produce algorithms that contain more parameters and produce more realistic results with much more economical ways and methods.
Why is it not enough?
The algorithm asks us for height and weight information but cannot interpret the regional weight distribution in our body. For this reason, the algorithm does not have an idea of how deep into the mattress the heavy areas such as shoulders and hips will sink and what the spine plane will be. The spinal plane is extremely important in terms of sleep hygiene. A mattress that cannot provide the correct support causes stretching of the spinal discs and ligaments throughout the night. Therefore, we wake up to a tired and painful morning.
The situation is not much different for the way of sleeping. You can give the algorithm the information that you are lying on your side, and the algorithm interprets this information and suggests a mattress. When we ask about the sleeping position information, the vast majority of people are not aware of exactly how they sleep because they are based solely on the position they are trying to sleep in. The vast majority of active sleepers sleep in mixed sleeping positions. These people, who frequently change positions during the night, are not even aware of the mobility. It is not possible for this highly personal information to be interpreted correctly by an algorithm.
What path should be followed?
Scientific data should take place before data based on informal observation in the algorithm setup. Taking advantage of actigraphy tests and sleep surveys are indispensable for a scientific algorithm. The material behavior of all mattresses offered for sale should be examined and the regional pressure should be determined with the help of the pressure mattress. For this, different morphotypes and age groups should contribute to the process as a subject. The number of subjects and body type diversity will increase the success rate.
Subjects should be allowed to try all mattresses for 5 hours, then 8 hours and be asked to score between 1 and 10 points after experience. The rate of movement during the night, the time between sleeping positions and the last sleeping position before waking should be reported. The position preferred by the subject involuntarily during sleep and the position chosen by the subject while awakening will be ready to be evaluated as an algorithm data.
Conditions such as the subjects’ feeling of comfort, pain conditions, opinions about sweating, morning energies, and perception of the hardness of the mattress should be learned by the survey method. It should also be classified in terms of age, body structure and sleeping pattern. The information collected should be real algorithm data and should be produced with scientific methodology.
Is the algorithm just enough?
The purpose of the algorithm is to reduce the number of options within the product range. Too many options will seriously confuse the customer. The purpose of the algorithm is not to pinpoint, but to contribute to the store presentation process. By eliminating inappropriate options, the algorithm gives the expert in the store qualified time for other options. During this time, the expert tries to understand which customer’s body structure the mattresses suggested by the algorithm fit perfectly. Therefore, the algorithm asks various questions and tries to satisfy the need.
The algorithm then allows the mattress to be tested and examines its effects on the spine axis. The straight spine line is an important sign of a homogeneous distribution of body weight. Immediately afterwards, the algorithm tries to understand the dermal pressure on the shoulder and hip by asking various questions. The mattress will begin to give an idea to the user after trying at least two different sleeping positions and a minimum of 15 minutes in total.
After evaluating sweating, allergic sensitivities and possible musculoskeletal problems, the person will have the right product for him. By eliminating the remote options of the algorithm, it is possible to get a good result thanks to the knowledge and experience of the expert advisor. The right product can be obtained by scientific production methods. The correct algorithm can be designed using data after human-product interaction. Consequently, the right institutions never compromise on these two factors.
Healthy Sleep.
Ersin Demir
Sleep Specialist / Trainer-Consultant / Sleep Ür. Researcher