In 1999, when NASA’s Mars Climate Orbiter skipped its target orbit and burned into the Martian spacecraft, the media had a great day for one reason: one group had used metric units in its calculation, the other, imperial. Navigation software that exchanged this information did not have a built-in unit testing system. So when one-party software produced data on state units rather than expected metrics, the spacecraft was set in the wrong direction. The result was a five-year loss and hundreds of millions of dollars in taxpayers’ money.
Twenty years later, similar problems persist. Researchers in all fields tend to think that their colleagues understand the details without clarifying them, and as a result, they err on the side of caution. Sometimes they leave it out altogether, provide many meanings or use simple units that have never been officially recognized.
People find it difficult to translate numbers with careless or non-existent units, and it becomes even more difficult when computers are involved. Most software packages, data management tools and programming languages ​​do not have built-in support for combining units and numerical data (other than the F # language). This means that information is stored and treated as ‘unified’ values. Behavioral processes that include bioscience and aerospace engineering have adopted unit representation agreements, such as the Integrated Unit Standards (UCUM) and Values, Units, Dimensions, and Types (QUDT) Ontology. But no technical definitions have been widely agreed on how to represent prices and related units other than confusing machines.
There have been a number of calls in recent years to create FAIR data sets (Available, Accessible, Interactive and Usable), and to ensure that open data is in line with the 5-star feed system proposed by World Wide Web founder Tim Berners-Lee, which aims to make them accessible, comfortable and organized. Many researchers are now committed to uploading data to free and open source databases.Tensions between units undermine these efforts. After all, many scientists spend more time confusing data than on research. If data cannot interact or be machine readable, information for individual researchers is blocked. The benefits of data sharing are diminishing.
Unless we take steps to ensure that the units of measurement are standardized for convenience, consistent data exchange, information will not be usable or, worse, misinterpreted. All the challenges of the world, from epidemics to climate change, require high quality data from all sectors, from international sources. Mistakes and lost opportunities will cost mankind hundreds of millions of dollars in a single shipwreck. In 1999, when NASA’s Mars Climate Orbiter skipped its target orbit and burned into the Martian spacecraft, the media had a great day for one reason: one group had used metric units in its calculation, the other, imperial. Navigation software that exchanged this information did not have a built-in unit testing system. So when one-party software produced data on state units rather than expected metric, the spacecraft was set in the wrong direction. The result was a five-year loss and hundreds of millions of dollars in taxpayers’ money.
Twenty years later, similar problems persist. Researchers in all fields tend to think that their colleagues understand the details without clarifying them, and as a result, they err on the side of caution. Sometimes they leave it out altogether, provide many meanings or use simple units that have never been officially recognized.
People find it difficult to translate numbers with careless or non-existent units, and it becomes even more difficult when computers are involved. Most software packages, data management tools and programming languages ​​do not have built-in support for combining units and numerical data (other than the F # language). This means that information is stored and treated as ‘unified’ values. Behavioral processes that include bioscience and aerospace engineering have adopted unit representation agreements, such as the Integrated Unit Standards (UCUM) and Values, Units, Dimensions, and Types (QUDT) Ontology. But no technical definitions have been widely agreed on how to represent prices and related units other than confusing machines.
There have been a number of calls in recent years to create FAIR data sets (Available, Accessible, Interactive and Usable), and to ensure that open data is in line with the 5-star feed system proposed by World Wide Web founder Tim Berners-Lee, which aims to make them accessible, comfortable and organized. Many researchers are now committed to uploading data to free and open source databases.
Tensions between units undermine these efforts. After all, many scientists spend more time confusing data than on research. If data cannot interact or be machine readable, information for individual researchers is blocked. The benefits of data sharing are diminishing.
Unless we take steps to ensure that the units of measurement are standardized for convenience, consistent data exchange, information will not be usable or, worse, misinterpreted. All the challenges of the world, from epidemics to climate change, require high quality data from all sectors, from international sources. Mistakes and lost opportunities will cost mankind hundreds of millions of dollars in a single shipwreck.
A world without unity
Many measurements are taken and reported outside units in the everyday world. Units are usually considered in a particular context. Take the temperature – in the ’20s’ is very cold in the United States, using Fahrenheit, but the cooler summer day in countries using Celsius. And cholesterol is measured in milligrams per deciliter or millimole liters, depending on the country. Skilled people can often determine what is meant by unadulterated numbers in scientific papers and data sets, but not always. The task of solving such problems is even more difficult for computers, which can often draw context and logical reasoning.
Some units mean different things in different situations. The C-capital calorie, used to describe food energy, is equal to 1 calorie – usually the amount of energy needed to heat a kilogram of water by 1 ° C at normal air pressure. Thus, calories and Calories vary by 1,000 factory, but the term cal (low-case c) is widely used in both. While the intended meaning may be obvious to someone interested in thermodynamics or the nutritional value of a hamburger, it is unclear to the computer. Similarly, the constant gravitational force G is often confused with g, local acceleration due to gravitational force, yet g is also used in grams. The meter is sometimes labeled M, which is also a mega element, and a unit of molarity. These principles and others cause computers to stumble.
Generally, the same values ​​are represented by different units. Melting, for example, is officially expressed as kilograms per liter (kg l – 1) or moles per cubic decimet (moldm – 3). These can be easily changed, but only if the units are properly marked. And sometimes the same unit is written in many ways. The microgram can be written as mcg, ug or µg. Speed ​​in meters per square second can be represented as m / s2, m / s ^ 2, m / s2 or m.s − 2. The typesetting principles use a range of character sets, italics, bold, slash, uppercase and subtitle. This is obvious to people, but it is not very consistent in that it can be faithfully studied by machines. There are too many units and too many variations that can be automatically analyzed or mapped all into a clear and usable representation.
Computer systems used for downloading and sharing data are not set up to assist. Take a simple example of Excel spreadsheets: the only unit that can be included in usable fields is a currency symbol. The association of a unit with a numerical value is left to unreasonable, invariant actions, such as a unit character unit given in the header row. That connection is easily broken when data is transferred or used in statistics.
Problem solving
There is much work to be done to address these issues. Many of the standards, principles and best practices of the units are readily available. The internationally recognized International System of Units (SI units) provides common names and representations of numerical values ​​and related units. Other international efforts have also achieved significant funding, for example through the International Organization for Standardization (ISO), the International Electrotechnical Commission (IEC) and the United Nations Economic Commission for Europe.
The production platform FAIR Digital Objects (FDO Forum) aims to improve the representation and transfer of scientific knowledge, including fully mechanized semantics. In fact, FAIR Digital Objects “integrates all the important information about a business into one place and creates a new kind of practical, sensible, and technologically permeable whole of life today”, according to the forum. But there is still much work to be done.
Since its launch, DRUM and Digital SI have been working to raise awareness and support efforts to improve collaboration with national and international organizations, including CIPM, the International Science Council, the Research Data Alliance and the GO FAIR Initiative.Another, most important goal has been taken by the high-performance SI Digital Task Team appointed by the Digital SI Expert Team: to create a robust, non-invasive data exchange framework based on SI units. This will help to solve long-standing problems in a strong way. For example, it may prevent the practice of representing certain units of number in many ways, in order to ensure that future plans do not perpetuate the problems plaguing the digital environment today. unions and associations, and the DRUM team are conducting a study on how the units are being used, the results of which will be reported later this year.
Entire scientific community needs to agree on a model
That report is intended to be a stepping stone. The entire scientific community needs to agree on a model that will represent numbers and units. This should include the official definitions applicable to the people and the processing of the machine. Databases that allow access to this information must be established. They should use service-focused infrastructure (such as websites and computer applications) to access information and modification of units. Editing areas, analytics software and data storage platforms should be the ‘information unit’.
DRUM can see this work, but it will not succeed without the extensive cooperation of all the many scientific and technological know-hows. Funding institutions and private companies should support this initiative, which is being carried out by volunteer groups like us. Providing even a small portion of the current R&D support to the workplace will bring broad, substantial benefits and enable national and international agreements to promote the use of clear, co-operative units.
Everyone agrees that comprehensible, useful data is at the heart of good science, and that information from various fields is needed to understand and correct global problems. Research programs do not meet those requirements. Time to make data and information easily accessible to machines and people.
READ ALSO : National Electron Tomography Workshop for Biological specimens: New Delhi
