A mini-engine harnesses noise to convert information into fuel

Schematic information engine. (a) The noisy detector measures the position there of the bead, located in fact at X. Ratchet based on (b) noisy position measurement there or (c) Bayesian position estimation _X (dotted blue circle). Credit: Physical examination letters (2022). DOI: 10.1103/PhysRevLett.129.130601

Too much background noise is usually guaranteed to disrupt work. But physicists have developed a micro-scale engine – made from a glass bead – that can not only resist the distracting influence of noise, but also harness it to operate efficiently. Their experience is reported in the journal Physical examination letters and was selected by the journal as a highlight of the research.

In everyday life, we know the machines and engines that consume fuel to move in a directed way and thus perform useful work. But things are more complicated in the microscopic world, where noise in the form of heat can easily sabotage things.

“Heat makes the components of small machines move all the time,” says lead author John Bechhoefer, a quantum physicist at Simon Fraser University in Burnaby, British Columbia, and a member of the Foundational Questions Institute, FQXi, a group thinking about physics. . So generally the effect of such thermal noise due to heat in the environment is to reduce the amount of useful work a small motor can do.

But there is a special family of microscopic machines, called “information engines,” that can actually harness noise to move in a directed way. An information engine works by measuring small movements caused by heat and using this information to selectively reinforce movements that go in the “right” direction, in the direction required by the machine.

“An information engine is a machine that converts information into work,” explains Bechhoefer.

Physicists and engineers are excited to build such small information mining engines to design new microscopic machines for nanotechnology applications. “There is great interest in taking inspiration from the biomolecular machines that nature has developed,” says co-author David Sivak, a physicist also at SFU. “Our work advances our understanding of how information can be used in such machines, pointing to possible uses for sustainable energy harvesting or more efficient computational storage and computation.”

“An information engine is a machine that converts information into work”, explains John Bechhoefer.

Bechhoefer, Sivak and their SFU colleagues Tushar Saha, Joseph Lucero and Jannik Ehrich built an information engine using a microscopic glass bead – the size of bacteria – suspended in water. The cord is held loosely in place by a laser beam which acts as a support under the beam. The molecules in the water gently jostle the pearl, due to natural thermal fluctuations in the liquid, and occasionally the pearl will be shaken.

Here’s the trick: when the team measures that the bead has moved against gravity, due to thermal fluctuations, it lifts the laser stand. In this higher position, the bead now has more stored energy, or gravitational potential energy, like a ball that is held in the air, ready to fall.

The team did not have to expend any work to lift the particle; this movement occurred naturally thanks to the jolts of water molecules. Thus, the engine converts the thermal heat of the water into stored gravitational potential energy using feedback on the bead’s motion to adjust the laser trap. “The decision whether or not to lift the trap, and if so by how much, depends on the information we gather about the position of the bead, which acts as the ‘fuel’ for the engine,” says lead author Saha.

This is how it works in principle, but implementing the strategy correctly is difficult if there is too much measurement noise, generated in the system from the brightness of the laser beam used to locate the bead. In such cases, the uncertainty in the position of the bead for each measurement can be greater than the bead movements produced by the shaking water molecules. “Measurement noise leads to erroneous feedback and thus degrades performance,” says Saha.

“Bayesian” information engine

Typical information engines use feedback algorithms that base decisions on the last measurement of bead position, but these decisions can be wrong when measurement errors are large. In their recent article, the team wanted to investigate if there was a way around this disruptive issue.

They developed a feedback algorithm that did not simply rely on a direct measurement of the bead’s last position – as this measurement could be inaccurate – but rather a more accurate measurement of the bead’s last position based on all measurements. previous ones. This filtering algorithm was thus able to take into account measurement errors in the development of its estimate, called “Bayesian estimate”.

“By combining many noisy measurements in a smart way involving a model of the bead’s dynamics, one can recover a more accurate estimate of the bead’s true position, greatly mitigating performance losses,” says Lucero.

In their new experiment, reported in Physical examination letters, the team demonstrated that an information engine that applies feedback based on these Bayesian estimates performs significantly better than conventional information engines, when measurement errors are large. In fact, most typical news engines will shut down if the measurement errors are too large.

“We were surprised to find that when the measurement errors exceed a critical threshold, the naïve engine can no longer function as a pure information engine: the best strategy is to simply give up and do nothing,” explains Ehrich. “In contrast, the Bayesian information engine is able to produce small positive work regardless of the amount of measurement error.”

There is a price to pay for the ability of the Bayesian information engine to extract energy even with large measurement errors. Since the Bayesian engine uses information from all previous measurements, it needs more storage capacity and involves more information processing.

“A trade-off arises because reducing measurement error increases the work extractable from the fluctuations, but also increases the information processing costs,” says Ehrich. The team thus found maximum efficiency at an intermediate level of measurement error, where they could achieve a good level of energy extraction, without requiring too much processing.

“There is great interest in taking inspiration from the biomolecular machines that nature has developed”, explains David Sivak.

The team is currently investigating how things might change if the noise that “powers” the engine comes from something other than heat. “We are preparing a paper that investigates how the optimal feedback strategy and performance change when fluctuations are no longer simply thermal,” Saha explains, “but also arise due to active energy consumption in the surrounding medium, such as c is the case in living cells”. .”