Skip to main content
Disciplines
Automotive Business IT + Informatics Construction + Real Estate Electrical Engineering + Electronics Energy + Sustainability Insurance + Risk Finance + Banking Management + Leadership Marketing + Sales Mechanical Engineering + Materials
Events
DE
EN
Books
Journals
Topic Page
Marketing
Start single access now
Access for companies
EXTENDED SEARCH
Log in
Top

2023 | Book

Introduction Read chapter Read first chapter

Information-Powered Engines

Author: Tushar Kanti Saha

Publisher: Springer Nature Switzerland

Book Series : Springer Theses

Part of: Springer Professional "Wirtschaft+Technik" , Springer Professional "Technik" , Springer Professional "Wirtschaft"

Login to get access
insite
SEARCH

About this book

This book presents the experimental development of an information-powered engine inspired by the famous thought experiment, Maxwell’s demon, to understand its potential to produce energy for practical purposes. The development of an engine based on Maxwell’s demon was for a long time inconceivable, but technological advances have led to novel investigations into theoretical and practical applications.

The built information engine consists of a micron-sized glass bead trapped in a tightly focused laser beam. It rectifies the bead's Brownian motion by controlling the laser's position and generates a unidirectional motion against gravity without doing any work, thus converting thermal heat into stored gravitational potential energy. A theoretical model based on a spring-mass system describes the engine's dynamics and was then used to find optimum parameters to improve the engine's performance. Experimentally implementing these optimization strategies led to engine output powers comparable to those measured in biological motors.

This book also highlights performance improvements made in the presence of measurement noise and presents important guiding principles to design information engines to operate in non-equilibrium environments. By focusing on practical applications, the book overall aims to broaden the scope of information-engine investigations.

Table of Contents

Frontmatter
Chapter 1. Introduction
Abstract
Engines have shaped the evolution of modern human society. We heavily depend on machines such as mobile phones, cars, and airplanes. Engines are machines that convert some form of energy into useful work. Some of the more common traditional engines include heat engines, hydroelectric power plants, and wind-powered turbines. Heat engines convert heat, generated from combustion of charcoal, diesel, etc., into mechanical motion. Hydroelectric power plants convert the potential energy of water, stored in a dam, into electricity by driving water turbines. Wind turbines work on a similar principle; they generate electricity from rotating blades, driven by wind energy.
Traditional engines were the backbone of the Industrial Revolution. They powered big factories that produced goods in large quantities and heavy vehicles that transported these goods. Most macroscopic engines operate at large energy scales and are usually not affected by the thermal fluctuations in their environment, which are of order \(k_{\mathrm {B}}T\) in energy. (Here, \(k_{\mathrm {B}}\) is Boltzmann’s constant and T is the temperature of the local environment.) By contrast, small molecular motors, which power cellular functions in living organisms, operate against fluctuating forces that are much larger than those produced by these motors. However, they have naturally evolved to work efficiently in such dissipative and noisy environments.
In Fig. 1.1, we show one way to classify the different kinds of engines. Note that the list is not complete, and there are other ways to classify engine types (for example, by the fuel source). But the outline serves to place the main focus of this thesis—energy-harvesting information engines—in the context of engines that have been built and studied.
Tushar Kanti Saha
Chapter 2. Theory Background
Abstract
A microscopic particle immersed in a fluid undergoes fluctuating motion in the absence of any visible forces acting on the particle. This phenomenon was first observed by Robert Brown in 1827 while looking at pollen grains in water under a microscope (Brown, Philos. Mag. 4:161–173, 1828; Li and Raizen, Ann. Phys. 525:281–295, 2013). The jittery motion of microscopic systems is now known as Brownian motion (Jacobs, Stochastic Processes for Physicists: Understanding Noisy Systems. Cambridge University Press, Cambridge, 2010, Sec. 5.1). The fluctuating motion of the particle arises from its collisions with the fluid molecules. During such collisions, the Brownian particle exchanges energy with the environment; these energy exchanges are at the heart of our information engine. In this chapter, we first discuss the equation of motion of a trapped Brownian particle, a micron-sized bead. Since we measure the position of the bead at discrete time intervals, we derive the discrete-time equation of motion describing the bead’s position from one measurement to the next. The motion of the bead arises from external forces, which could be either deterministic or stochastic. The work done by the external forces is measured by using techniques from stochastic thermodynamics. We then discuss the discrete-time energy estimates that are used in this thesis to measure the energy flows in experiments. Each realization of a stochastic-system evolution follows a different trajectory. These trajectories are best described by a probability density function. Finally, we study the time-dependent evolution of the probability density function of a trapped bead using the Fokker-Planck equation.
Tushar Kanti Saha
Chapter 3. Experimental Apparatus
Abstract
In this chapter, we discuss the experimental apparatus that is used to build, control, and characterize the information engine. We discuss the calibration routines that convert the (nonlinear) voltage signal from the detector into a position of the bead in physical units. Then we discuss a method to measure the optical trap stiffness, bead’s diffusion constant, and the measurement noise of the position detector. We also present a method we developed to control the strength of the measurement noise. Then we discuss an optimal-filtering technique that improves the estimate of the bead’s position in the presence of measurement noise. We implement the real-time position estimate and benchmark its performance. Next, we modify the experimental apparatus to integrate a dual optical trap system and trap one bead in each trap. Finally, we modify the sample chamber design to introduce electrodes that can apply external forces on the bead.
Tushar Kanti Saha
Chapter 4. High-Performance Information Engine
Abstract
In this chapter, we create and study the performance of a useful information engine that not only extracts energy from heat but also stores energy by raising a weight, as initially imagined by Szilard. The “fuel” for the motor is the information gathered from favourable system fluctuations.
In our study of this information engine, we focus on understanding and then optimizing its performance: How much can it lift? How fast can it go? More precisely, what is the upper bound to the rate of gravitational energy storage and to the directed velocity? We reason that the value of the function of a motor can greatly exceed the cost of running it. For example, in biological applications such as chemotaxis, the metabolic costs of running cellular machinery (including information-processing costs) are usually unimportant compared to the benefit gained by the ability to move toward a new food source or away from a predator.
Tushar Kanti Saha
Chapter 5. Trajectory Control Using an Information Engine
Abstract
In Chap. 4, we presented an information engine that could raise an optically trapped bead against gravity and store gravitational potential energy. We found that the average velocity of the bead depends on experimental parameters such as trap stiffness and bead diameter. Here we operate the information engine to convert thermal fluctuations into directed motion, without storing any gravitational energy. We term such an engine a transporter engine. Although the ensemble moves at an average velocity, individual trajectories have velocities that fluctuate independently about the average. As a consequence, the algorithm described in Chap. 4 cannot control individual trajectory paths. In this chapter, I explore a feedback algorithm that can make the bead trajectory track a desired path, with zero input trap work. I characterize the performance of the feedback algorithm, drawing on ideas from control theory. I find that the frequency range over which the heuristic feedback algorithm can track is comparable to the corner frequency of the trap.
Tushar Kanti Saha
Chapter 6. Bayesian Information Engine
Abstract
In this chapter, we discuss an experimental realization of an optimal Bayesian information engine that retains the relevant memory of all past measurements in a single summary statistic. Using the extra information from past measurements and correctly compensating for delays in the feedback loop via predictive estimates, we extract and store significant amounts of energy, even in the presence of high measurement noise.
Our implementation of the Bayesian filter uses the optimal affine feedback control algorithm, at optimal experimental parameters, to maximize the engine’s rate of gravitational-energy storage. The relevant information from past observations is used to minimize the uncertainty in the bead’s position. This Bayesian information engine extracts energy even at low signal-to-noise ratio (SNR), avoiding the phase transition in the naive information engine that leads to zero output. Under any conditions, this engine extracts at least as much work as the naive engine and, indeed, reaches the upper bound on the performance of Gaussian information engines.
Tushar Kanti Saha
Chapter 7. Information Engine in a Nonequilibrium Bath
Abstract
In this Chapter, we present an information engine that is coupled to a medium that is at an equilibrium temperature T but also exhibits external noise beyond what is expected from the thermal fluctuations. Such a bath acts as an energy source with contributions from both thermal and nonequilibrium (external) fluctuations. We show that an information engine can extract significantly more energy from this type of nonequilibrium bath than it can from an equilibrium thermal bath—up to ten times more power than from a bath at room temperature, for the nonequilibrium baths that we are able to create.
Tushar Kanti Saha
Chapter 8. Identifying Information Engines
Abstract
In this Chapter, I seek methods to distinguish the physical aspects of the information engine from a power-stroke (heat) engine. In general, the operation of the molecular motor could be very complicated. A simple power-stroke mechanism, used to model molecular motors, involves relaxation of a constrained spring that is triggered by a chemical reaction. Here the term “power stroke” is used to refer to a slightly different mechanism, one where the spring applies a constant (average) force: The motor simply moves the trap centre at constant speed, so that the bead on average moves at the same speed. To isolate the differences between the two kinds of engines, I use the same experimental apparatus. I measure the energies flowing in the power-stroke engine and find that a power-stroke engine dissipates heat into the medium because of friction between the bead and the viscous medium—water. I also consider a “drift-diffusion” motor, which adds random forcing to the steady power-stroke motion. Again, heat is dissipated into the medium. By contrast, a translating bead powered by an information engine does not dissipate any heat; rather, it extracts heat from the medium and converts it into useful energy. That heat is absorbed by an information engine while it translates through a medium distinguishes it from other power-stroke engines.
Tushar Kanti Saha
Chapter 9. Conclusion
Abstract
Information engines operate by converting heat in the surrounding medium into energy. They have refined our understanding of the second law of thermodynamics and our understanding of information as a physical quantity. Previous experimental studies were directed towards showing that these engines operate in a way that is consistent with the second law of thermodynamics, even though they may superficially seem to violate it. However, in this thesis, our goal has been to understand and optimize the performance of information engines. Information engines differ from molecular motors and heat engines: They use feedback to extract energy directly from the surrounding environment, paying only the cost of running the measurement and control apparatus. By contrast, molecular motors and heat engines use the energy from another source of fuel to operate.
Tushar Kanti Saha
Backmatter
Metadata
Title
Information-Powered Engines
Author
Tushar Kanti Saha
Copyright Year
2023
Electronic ISBN
978-3-031-49121-4
Print ISBN
978-3-031-49120-7
DOI
https://doi.org/10.1007/978-3-031-49121-4

Premium Partners

    Image Credits