Sunday, September 29, 2019

Energy, Trade-offs, and Automation


     It takes energy to do things. The verb "do" pretty much says that effort is needed. To "think" may not require direct physical work -- but, in order to be able to think, it is necessary that the body and brain (and the "mind" -- whatever that truly is) function -- which requires a lot of physical work by the parts of the body.
     The basic source of energy is, of course, the sun. Energy from the sun allows both physical and biological processes to continue. Plants grow and the energy is stored. Animals eat plants and other animals. As something is consumed, it is used as energy for further activity. From a physics point of view, matter is exchangeable with energy but, within biological systems such as a stalk of wheat, a butterfly, or a human, it is a release of stored energy via biochemical processes that keep one foot going after another.
     So, we "do" things. We walk. We run. We hunt. We eat. The first level was that all the energy was supplied directly from our bodies. The next level has us working with tools -- tools that allow use of other types of stored energy (kinetic, gravitational, wind, tidal, ...) Our use of energy allows us to make use of other forms of energy. This is called "leveraging" (probably not coincidentally similar to one of the basic tools -- the lever).
     The next level is to build other devices that can make use of stored energy autonomously (without active control and manipulation by humans or other animals). So, we make use of a wind-up or an electrically-powered clock. We use appliances that can be started and then will continue to work on their own. A train uses stores of fuel to move lots of other things. On a track, they can continue on their own (although normally a human is required to start them, stop them, or change tracks).
     There was once a statement (I cannot remember, or find, the source) that it was wise to consider the time needed to walk to a destination versus the amount of time and work needed to be able to buy a ticket on a train to a destination. Of course, such a comparison requires the need to value your time. If a ticket cost $100 and it takes you 20 hours at $5/hour to earn the money but it only takes you 15 hours to walk the distance then walking is less expensive. Of course, if you earn $10/hour then it is advisable to take the train.
     This type of value/cost comparison is relevant in the case of automation. If a human (or group of humans) with tools can create a shoe in 1 hour at $10/hour, then the base cost (plus materials and tools and so forth) is $10. If a machine can make 1 shoe a minute and it can operate 24 hours a day and it costs $300,000 to design and purchase it, with $500/month maintenance then it would pay for itself in less than six months. Obviously, if the humans can create a shoe in 1 hour at $1/hour, it would take more a little less than five years.
     But, let us return to energy. An automated shoe maker will need energy. If that is easily available and cheap then the formula just mentioned will be a guiding force for what to do in a company. What if it isn't? Then that human who converts their own energy has a clear claim to the work. So, we now have three factors, cost of tools/machinery cost of labor, and availability and cost of energy to run the machinery.
     As discussed in previous blogs, capital/money is a form of stored labor -- or stored energy. A final factor in calculation involving the profitability is the locking up of capital. A company that has 6 shoe making machines would have $1.8 million tied up in the machinery. What else could they be doing with that money? Perhaps they could make more profit using that money elsewhere?
     We now have four factors to determine the attractiveness of outsourcing or automation: cost of labor, cost of tools/machinery, cost of energy, and loss of flexibility via the locking up of capital into fixed equipment. Automation requires a consistent source of energy. Without that, it is best to stay with manual labor (because "cost of energy" becomes impractical to overcome).
     The economic practicality of automation can be defined in a straight-forward manner; accountants can determine best practices rather easily. The social implications are not so clear-cut. Automation shifts, and reduces, the needs from the labor force. More education is needed for design and maintenance of the automation and fewer people in those positions. This implies retraining of those who will take those positions as well as further education and training of those displaced from the industry. In addition, concentration of capital (via the owners of the automation and automated factories) continues concentration of wealth and income for which there needs to be methods of distribution.
   

Interrupt Driven: Design and Alternatives

       It should not be surprising that there are many aspects of computer architecture which mirror how humans think and behave. Humans des...