Designing Sensor Systems for Big Machines

Informatics: Noisy data is useful when processed in millions

Current MEMS sensors are very noisy. But it is alright because the current processor can process data in an astonishing speed of 90 million data points per seconds, plenty fast enough to do various filtering and averaging to produce a precise and accurate readings. Coupling the two together results in a cheap, sensitive and reliable sensor systems that have been breaking grounds in almost every fields of works.

Stability is the perception of precision and accuracy combined

Good sensors are sensitive, causing the data to fluctuate in the matter of thousands of times per second. When the numbers jump about, while it is more precise and accurately show the sensor’s true capability, it gives psychological doubt in the human’s eyes. We humans do not appreciate raw accurate truth as much as stability and certainty of a value. Secondly, we also want the human operator to focus their attention on the surroundings, not only on the readings when operating the big machineries like cranes and excavators. Thus, when designing the display, it was found that 2 decimal points, and freezing the data when the value fluctuates within 0.1 of the previous values are the most comfortable to the operator’s eyes. 

Machine-to-machine usability

It is a common scenario that nowadays we are designing a product with machines and computers as the user. The human operators are just an observer of the end-result of both the machine and the sensor system. Thus, the user-study requirement consists of the machine’s dimensions in replacements of ergonomics, and datasheet and operating specs instead of user requirements. As engineers, this has been a common practice, but treating machines as a user brings many other interesting perspectives, especially since in the very near future, these machines will be as complex as a human where the datasheet won’t be enough to contain all the user requirements, and the designers would really need to understand the robots and “interview” them by using them for a while in order to design a product that works best for them.

Previous
Previous

3D Printing for Home Users

Next
Next

Design Requirements