ERICSSON ZERO ENERGY PROTOTYPE

Hi All

The following is for long term genuine investors and those who are new to Brainchip and wonder if the technology is revolutionary and has the potential to become ubiquitous.

I will not be expressing any opinions about the following material and will provide the references to the complete articles so you can ensure that

the extracts I have included have been taken in proper context.

The one thing I will say is that I posted a question at the recent investor event asking if Brainchip was aware of the published paper by Ericsson wherein they had developed a successful prototype of a Zeroenergy device using AKIDA technology but unfortunately my question was not addressed:

Extract1.

In 2022 Ericsson published the following paper:

6G networksand devices will increasingly rely on AI, requiring energy-efficient computing.Here at Ericsson, we recognize neuromorphic computing as a promising paradigmfor this. We havejoined forces with researchers at MIT who in their turn recognizenext-generation lithium-based oxides as key building blocks of neuromorphiccomputing.

https://www.ericsson.com/en/blog/2022/6/lithionic-memristors-future-neuromorphic-computing

Extract2.

Ericsson published the following paper:

“This will be complemented by zero-energy devices, a class of devicesharvesting energy from the surroundings and providing input to digital twins.AI/ML will also play an important role in the fully data-driven architecture of6G and the intelligent network platform of the future.”

https://www.ericsson.com/en/5g/5g-networks/5g-advanced

Extract3.

Ericsson published the following paper:

https://www.ericsson.com/en/6g

“Extreme energy performance

High network energy performance was an important requirement in thedevelopment of 5G, and it will be even more important for future wirelessaccess solutions. It is critical that the expected massive increase in trafficwill not lead to a corresponding increase in energy usage. An acceleration intraffic should not mean accelerated energy usage. Also, the energy usage shouldbe close to zero when there is no traffic within a node.

“Resilience, security andtrustworthiness

As wireless networksincreasingly become critical components of society, resilience and securitycapabilities are crucial. The networks must be able to provide service whenpart of the infrastructure is disabled due to natural disasters, localdisturbances, or societal breakdowns, and they must offer robust resistanceagainst deliberate malicious attacks. In terms of trustworthiness, the networksshould be able to leverage new confidential computing technologies, improveservice availability, and provide enhanced security identities and protocolswith end-to-end assurance.”

“Dependable compute and AIintegration

6G networks will need thecapabilities of dependable compute and AI integration, infrastructure enablingdistributed applications and network functions to be swiftly developed anddeployed, and services for data and compute acceleration, which can be deliveredthroughout the network with performance guarantees.”

Within the linked document above there are other links and in one youare taken to North America and discover the Next G Allianz:

6G Technologies

July 11,2022

Leadership

• Tingfang Ji (Qualcomm), Chair
• Sharad Sambhwani (Apple) , Vice Chair
• Stephen Hayes (Ericsson), Vice Chair
• David Young (ATIS)

© 2024 ATIS All rights reserved. | Privacy Policy

Extract 4.

In 2023 six researchers from Ericsson released the following research paper:

Towards 6G Zero-Energy Internet of Things: Standards,Trends, and Recent Results

Talha Khan, Sandeep Narayanan Kadan Veedu, András Rácz,Mehrnaz Afshang, Andreas Höglund, Johan Bergman Ericsson {talha.khan,sandeep.narayanan.kadan.veedu, andras.racz, mehrnaz.afshang, andreas.hoglund, johan.bergman}@ericsson.com

Abstract—6G presents new opportunities toenrich the cellular ecosystem by introducing battery-less Zeroenergy Internetof Things (ZE-IoT) devices, thus unleashing an era of massive, sustainable, andsmart connectivity. This explains the increased interest in ZEIoT in academiaand industry. The road to a 6G future empowered by ZE-IoT entails cohesiveefforts in the realm of standardization, academic research, and industrialtrials, which are synergistic with the anticipated market demand and thedominant technology direction. In this article, we provide a holistic view of a6G ZE-IoT future informed by the ongoing standardization activities in the 3rdgeneration partnership project (3GPP) for ZE-IoT, the role of the emergingtechnology trends such as digital twins and artificial intelligence, and thetechnical challenges in integrating ZE-IoT into the cellular ecosystem.Finally, we present some recent research results to address some of thediscussed challenges.

Keywords—3GPP, NR, IoT, Energy harvesting,Backscatter communication, Ambient-IoT, A-IoT, 6G, Green communications, OFDM,Zero-energy, ZE-IoT…

A. Usecases and deployment scenarios

Based on their functionality andapplication, the use cases for ZE-IoT were grouped as follows in the RAN studyitem:

 Inventory: Examples includeautomated warehousing, end-to-end logistics, automated supply chaindistribution, etc.

Sensors: Examples include smart homes, smart agriculture, smart grids, etc.

Positioning: Examples include finding remote lost items, positioning inshopping centers, location services, etc.

 Command: Examples include deviceactivation and deactivation, elderly health care, electronic shelf label, etc….

B. A Communication stack prototype for AI-enabled ZE-IoT devices

To demonstratethe feasibility of low-energy AI and low-energy communication in a ZE-IoTdevice, we have built an example use case, as illustrated in Figure 4. TheZE-IoT device consists of a low-power camera, a neuromorphic AI chip (theAkida neural chip from BrainChip), a low-powerradio and a solar panel.

The application assumes that the cameratakes a picture (e.g., when triggered by a motion sensor), runs a neuralnetwork to create the neural embedding of the image (i.e., extracts the neuralfeatures from the image), and sends the neural embedding vector via a customradio stack tailored for AI data that implements approximate and intermittentcommunication. The use case specific AIlogic is hosted in the network which implements the final layers of imagerecognition. It can be customized for object, face or gesture recognition.Consequently, the AI logic in the sensor device can be use case agnostic,allowing considerable flexibility to introduce new use cases by adding new AIlogic on the network side. Figure 4: AI-enabledZE-IoT prototype use case. The radio link includes a custom data encoding,where instead of using binary encoding for the vector elements and sending thedigital data with error correction encoding, we first create pseudo randomlinear projections of the embedding vector from the Ndimensional space tosingle-dimension and send these projected values on the radio link as digitallyencoded and modulated data or with a quasi-analogous modulation. Formally, [ ]=[]∗[ ] (1) where isthe raw vector containing the l projections, isthe embedding vector to be transmitted and contains the first l linear projection codewords. Note that is not transmitted as it can be regenerated at the receiver from the same seedas used in the transmitter. The index l continuously increases with everytransmission. The receiver attempts to obtain bysolving Eq. (1) based on the received and the known . The proposed encoding enables approximatecommunication as each transmission from the ZE-IoT sensor includes informationabout the entire embedding vector which can be reconstructed at the receivermore accurately with the reception of every new transmission. Even if theZE-IoT device has energy for only a single transmission, it can stillcontribute to the reception of the embedding vector. In Figure 5, we plot the energy profile ofthe ZE-IoT prototype device to illustrate the stored and spent energy versustime. When the harvested energy reaches the amount required for the nextprocessing stage (e.g., camera capture, AI inference or radio transmission),the task execution depletes the stored energy, which is again collected by thesolar harvester. Figure 5: Energycollected and spent.

The average values for energyconsumption and harvesting are as follows: (1) the AI inference on the Akidachip consumes ~8 mJ/image inference with 480k neural network parameters and4-bit quantization of the weights, (2) one radio transmission of a 20 bytesframe consumes ~5 mJ and requires ~4 ms transmission time, and (3) the solarharvester produces ~1.5 mW or equivalently 4.5 mJ/3 s power under typicalindoor lighting.

V.CONCLUSIONS

In this article, we have discussedvarious aspects of the 6G cellular ZE-IoT technology. We have comprehensivelyreviewed the standardization efforts for ZE-IoT including use cases, devicecategories, Time [sec] Energy stored [mJ] 5 10 15 20 2 4 8 Energy for the cameraEnergy for AI inference 1st transmission2nd transmission design targets andfuture roadmap. We have also identified the role of emerging technology trendssuch as digital twins, AI and smart textiles in facilitating the mass adoptionof ZE-IoT and vice versa. Moreover, we have provided novel research results toaddress some of the challenges facing ZE-IoT. For instance, to demonstrate thefeasibility of AI-enabled ZE-IoT, we have developed a prototype of asolar-powered AI enabled ZE-IoT camera device with neuromorphic computing. Inaddition, we have also emphasized the need of an OFDM-compatible physical layerdesign for ZE-IoT. To this end, we have devised techniques for OFDM-compatiblebackscatter communication for passive ZE-IoT devices and for OFDM-compatibleOOK for active ZE-IoT devices. There areseveral opportunities for future research. One promising direction is to designOFDM-compatible communication techniques for ZE-IoT. Another possibility is todrive empirical research at the intersection of ZE-IoT connectivity anddisruptive technology trends such as AI. The conditions are ripe for cateringto the ZE-IoT use cases within a cellular ecosystem.

https://www.techrxiv.org/doi/pdf/10.36227/techrxiv.170327143.35314208

Extract 5.

Finally Ericsson published this article:

Zero-energydevices and 6G networks

In less than a decade, oursocieties will function and evolve based on real-time data from trillions ofsimple, low-cost, zero-energy sensors distributed. We refer to this new class of devices as zero-energy devices and theyare part of our vision for future 6G systems.

Compared tosmart phones or mobile broadband devices, many future connected sensors wouldbe extremely low cost, low performant, and consume just a small fraction of theenergy. It must be possible to mass deploy such devices in an affordable andsustainable way, and the 6G network needs to support that.

Limitedenergy consumption would mean being in a deep sleep for long periods of time.When activated, they would then only have the capability to transmit data for abrief period before requiring further energy harvesting.

Talkingabout Physical world only makes sense in relation to an existence of a digitalworld. But a digital world has not been mentioned here (yet), so doesn’t fit.

https://www.ericsson.com/en/blog/2023/5/zero-energy-devices-sensor-driven-world

I will as I say not be expressing any opinions so if you are interested in the above further you will need to do your own research.

Fact Finder