Interoperability

Interoperability is explained in great detail in chapter three. In a lack of words the text defines that interoperability is a conversion of data from a unique system to another unique system. A personal example for me would be: In my high school drafting class I designed a house using RoboCad and several years later transferred that information to AutoCad. The objective was to convert a set of 2D drawings from RoboCad to a single 3D drawing in AutoCad. Since there is no computer interoperability between the respective software products, then the mode of interoperability was my own labor of redrawing the 3D version in AutoCad from the printed sketches in Robocad. This mode of interoperability is severely inefficient.  

The text stresses that interoperability is of paramount importance in building systems because of the many disciplines it incorporates. We as a building society should focus on how well we can transfer data from one aspect of the building phase to another. Cost to design to construction to occupancy is an example.

So how can we improve this exchange of data between hardware and software? The text elaborates that in order to achieve more robust data exchanges it is imperative to create a standard in which the data is exchanged. In effect this standard who figuratively funnel multiple data sets into a single recognizable solution.

There are problems that arise from creating a standard of data exchanges. Fellow classmate David Morrison hits the nail on the head with his explanation. David Says’ “A company that does not provide as much interoperability is typically one who wants its standard / program to be king above the rest. Whereas a company that is looking for high interoperability would most likely be a company that wants its product to infiltrate the market.” Essentially if a common standard is the key to interoperability then a single entity must design the system. If competing entities are designing for their best interest then it defeats the purpose of interoperability.

An interesting dilemma arises from the concept of interoperability. In essence the only possible way to achieve 100% efficiency of data transfer is to eliminate the transfer altogether. This would mean that competition must be eliminated. For example we would eliminate the use of different BIM products and use a one single product. If Revit were the product we choose as a society then Autocad and other modeling tools would be abolished.  But since competition is imperative for market survival the interoperability is a science that must be researched and improved.  Prof. Mitchell hinted at the idea that the future will challenge our interoperability prowess when multiple systems (such as robots) will have to co-exist in a societal fashion.

  SOURCE: 

BIM Handbook: A Guide to Building Information Modeling for Owners, Managers, Designers, Engineers, and Contractors.  Chuck Eastman, Paul Teicholz, Rafael Sacks and Kathleen Liston  Copyright © 2008 John Wiley & Sons, Inc.

Sensor Capabilities

Sensors have been around for a long time. As David Morrison pointed out, thermostats are a primitive relative of modern-day sensors. So how have sensors evolved up until the present day? Well it basically boils down to two attributes. First off our sensors, in a lack of works, have become more sensitive.  The precision and accuracy have become more acute in pin-pointing the true values of data, thus giving us better feedback. Secondly, as sensor technology progressed, our sensors have now been able to perform actions. It is expected in the future that sensors can perform daily activities for humans without any assistance or user feedback.

An example would be the self-driving car. It is expected that by 2040 that self-driving cars will make-up 75% of the world-wide vehicle fleet.  Major manufactures of cars, such as Toyota and Cadillac, have claimed to have developed a car that can safely maneuver through traffic.

Dr. Azim Eskandarian director for the Center of Intelligent Systems at George Washington University explains that the technology we have today is sufficient enough to introduce self-driving cars into real world traffic. The sensor capabilities have been tested to be more reactive to changes in traffic flow than their human counterparts. Dr.  Eskandarian then continues to explain that the obstacles are not the efficiency and precision of the sensor technology that holds back production, but rather the cost. He hints that “reliability and robustness” are what is holding us back.  So in other words when the cost of sensors start to decrease an increase in integration of self-driving cars will be evident.

So what’s the future? It is estimated that sensor on cars will improve highway efficiency by a factor of three. This will reduce infrastructure costs because the renovation and widening of highways will not be needed. Cars will be able to travel closer together and essentially double highway capacities because of sensors and computer precision.

Examples of sensors used in self-driving cars are radar and ultrasonic-sensor technology. Typically present-day prototypes are equipped with ultrasonic-sensor technology. However the future hints that self-driving cars will primarily rely on radar.

To sum it up, it is believed that present-day technology is sufficient for self-driving automobiles but the cost is just not there yet. As Elda Cifligu opined in her blog, the bulk of future research in sensor technology with be how we can make then cheaper. Only when the price is right for self-driving cars we will see them in everyday traffic.

Sources:

http://www.forbes.com/sites/eco-nomics/2012/09/25/self-driving-cars-will-take-over-by-2040/

http://www.forbes.com/sites/timothylee/2013/01/11/high-sensor-costs-wont-stand-in-the-way-of-the-self-driving-future/

http://www.edn.com/design/automotive/4368069/Automobile-sensors-may-usher-in-self-driving-cars