申请试用
HOT
登录
注册
 
VoltDB-TechTrends-Is-your-intelligence-on-the-edge
0 点赞
0 收藏
0下载
VOLTDB
/
发布于
/
282
人观看

物联网边缘计算。数字化转型。 5G。看来我们所听到的这些术语已有一段时间了。尽管公司对于如何运行以及如何将关键业务应用程序货币化的各种方法,都有自己的强大蓝图,但是它们自然会重叠,很快就会成为下一代的数据管理。
当然,画面仍然模糊不清:就像通过不对焦的相机镜头看到的一样。但一旦聚焦了该视图,就像打耳光一样。您突然清醒,突然意识到这段时间一直在发生什么,但是现在为时已晚:干扰者已经采取行动,您正在被干扰,您觉得现在应该预料到了许多年前看似明显的巨变。
因此,企业需要主动考虑而不是被动地考虑其数据体系结构。一旦您不得不对某些事情做出反应,就已经错过了机会,并且已经损失了很多金钱,时间和资源。另一方面,如果您始终领先一步,即使那一步的方向不正确,在变革浪潮席卷之前,您仍然有时间进行修正。
有鉴于此,VanillaPlus最新关于“边缘智能的技术趋势报告”为您提供了一个有见地,且可能改变游戏规则的机会,了解最新的数据管理趋势(尤其是边缘计算)将能迅速破坏您的技术堆栈以及竞争对手的技术堆栈。

展开查看详情

1.VanillaPlus ISSN 2634-9116 IS YOUR INTELLIGENCE ON THE EDGE? Report UK£99 but FREE for anyone that registers Vol. 2, No. 3 to IoT Global Network www.iotglobalnetwork.com SPONSORED BY Tech Trends – your new inside track to the topics that really matter.

2.NE W R EP OR O T GETTINNG TO STAND N ALONE 5G 5 Download thee new feature repor t where T-Mobile, Ericsson, Rakuten Mobile, Vodafon ne U.K . and VoltDB discuss real eal-ttime applications andd new ser vice revenue oppo or tunities in the 5G era. DOWNLLOAD THE REPORT on, please visit VOLTD T B.COM

3. Contents Editor’s overview Low latency decisions Edge analysis 20 Pandemic pressures organisations to push intelligence to the edge George Malim examines the latest analyst 04 predictions for the increased uptake of MEC Data lakes and streams give life to and edge intelligence intelligence at the edge Data, communications and computing 16 Edge ROI infrastructure are coming together but Why low latency is vital for edge intelligence at the edge relies on new tools success and processes, writes George Malim 5G, edge computing, digital transformation, digital twins, machine learning and AI look Analyst report like buzzword bingo but there are rational connections between each, writes Dheeraj Remella 06 Securing the edge Edge computing will be critical for digital transformation Edge computing is a key enabler for solutions that utilise emerging technologies 22 MEC proves the need for speed such as AI, 5G, cloud and IoT, and it has While there are sensible, industrial cost become critical to enterprises, writes Jim justifications for MEC, Nick Booth warns Morrish, a founding partner at Transforma against being distracted by fanciful Insights projections Sponsored interview About VoltDB VoltDB enables global organisations to leverage emerging 5G latency standards to power new revenue opportunities, transform their business and operational support system, or develop strategic integrations for their enterprise customers. The platform instantly derives value 18 from anomalous events captured across multiple streams of fast data, delivering precise Can it be safer on the edge? decisions in less than 10 milliseconds that 12 Antony Savvas explains why the edge is turning out to be one of the safer locations directly influence in-the-moment 5G and MEC encourage edge for hosting processing power and analysing monetization, prevent digital fraud, and support intelligence to become pervasive digital transformation initiatives. VoltDB is data VoltDB’s Dheeraj Remella tells George Malim purpose-built to address application-specific how the challenges of edge intelligence are scale and latency challenges and augment being addressed and why there will be no previous big data and messaging investments going back to centralised, remote to enable businesses to evolve from big data architectures analytics to fast data decisions. www.voltdb.com VanillaPlus Tech Trends covers technological & Managing editor: Editorial director & publisher: Business development: Published by: business developments for businesses enabled by the George Malim Jeremy Cowan Cherisse Jameson WeKnow Media Ltd. Internet of Things (IoT). © Copyright WKM Ltd. All Tel: +44 (0) 1225 319566 Tel: +44 (0) 1420 588638 Tel: +44 7950 279368 Suite 138, 80 Churchill Square, rights reserved. No part of this publication may be g.malim@wkm-global.com j.cowan@wkm-global.com c.jameson@wkm-global.com Kings Hill, West Malling, Kent copied, stored, published or in any way reproduced ME19 4YU, UK without the prior written consent of the Publisher. Digital services director: Designer: Tel: +44 (0) 1732 807410 Nathalie Millar Jason Appleby Tel: +44 (0) 1732 808690 Ark Design Consultancy Ltd ISSN 2634-9116 n.millar@wkm-global.com Tel: +44 (0) 1787 881623 VanillaPlus Edge Intelligence I Vol. 2, No.3 I 03

4. Editor’s overview DATA LAKES AND STREAMS GIVE LIFE TO INTELLIGENCE AT THE EDGE Data, communications and computing infrastructure are all coming together to enable intelligence at the edge so the processes, analytics, management and security are the next step, writes George Malim. Multi-access edge computing (MEC) has existed since server has come down, making various applications 2014 and has often been referred to as mobile edge practical. However, for many, latency to distant data computing. That tag is fine for wirelessly connected edge centres is still an issue and it makes more sense to devices but increasingly, edge devices can be supported process or analyse inputs and utilise edge intelligence to by fixed connectivity hence the change to multi-access. provide actionable insights. The edge device therefore This shift indicates a growing maturity and becomes not only a sensor collecting information but understanding that it’s not just mobile devices that can also something that can take immediate action according benefit from intelligence at the edge. to that data. The connectivity is then utilised to update the centralised record and With lower latency networks, enabled by 5G, LTE- enable functions such as Advanced and fixed fibre connectivity, the time it takes digital twins. to traverse the network from the edge to a centralised VanillaPlus 04 I Edge Intelligence I Vol. 2, No.3

5. Editor’s overview George Malim Tech Trends Where is the edge? automation, autonomous vehicles, Traditionally, the term edge refers to the end of a managed logistics and many network within 30 miles of an endpoint, such as a other applications. connected device or sensor gateway. However, this The deployments are definition is eroding as the edge is now often closer than so varied that it looks that. The edge could be in a vehicle or on the device itself like there’s a use case so it’s accurate to say the edge is moving further from the for virtually any centre and closer to the user and the connected devices organisation that has themselves. multiple sites or a fleet of devices in Hyperscale service providers, communications service deployment. providers (CSPs) and IT companies are all trying to deliver their definition of edge computing but confining However, the projected definitions are not helpful. What’s really needed is an open benefits will not understanding that the edge is what an organisation necessarily come wants it to be. For some, that will be storage and compute easily. Edge resources at the cellular network edge while for others, the intelligence presents a substantial shift from the cloud device itself will contain processing, storage and computing infrastructure of the previous generation. The analytical power to perform edge intelligence at the speed, while one of the main advantages, is also a massive device level. In contrast to traditional hub and spoke source of disruption because being able to perform tasks architectures in which edge data flows up and down a in near real-time transforms business processes which spoke to a centralised hub, MEC utilises a distributed cloud therefore have to adapt to become used to the new rate of computing model that enables processing and storage of progress. data at the network edge. These typically smaller-scale resources are closer to the devices and the end user and Speed also distorts perceptions and expectations and enable a cost-efficient, optimised network architecture. some of these may become unrealistic or excessively Don’t forget that traditional data centres and cloud expensive to support. For example, an organisation that computing has never been free so often MEC is not an relies on very high speed network connectivity to enable additional but an alternative cost and one that can be its application may find that network latency means it offset by the business benefits of low latency or the cannot meet its customer’s service level or that decisions localised processing of large volumes of data that are made at the edge are outdated because the data being critical to high-quality services. utilised is stale. With this edge infrastructure in place, the next step is to Edge intelligence is also a complex environment involving extract intelligence from it. The intelligent edge combines multiple technology types that all require integration, connectivity, computing power, artificial intelligence (AI) security and monitoring. There is substantial risk that this and data analytics so data can be acted upon more quickly complexity can mushroom making it hard to identify root and much closer to the point at which it is captured. causes of issues or expensive to monitor effectively. These As organisations generate ever-greater amounts of data are the downsides and are common to large-scale they are engaging in more complex operations and the immature technology roll-outs and the technology markets they operate in are transforming at the same industry is working hard to address them. However, it time. Real-time data places more demands on businesses remains early days and some of the first edge intelligence to sense more and be responsive to the sensor information deployments are hampered by complex architectures, they collect. Using data to gain insights quickly and poor integration between different systems and lack of effectively drives operational efficiencies and delivers availability of low latency connectivity at various competitive advantages and this underpins the huge locations. expected growth in MEC and edge intelligence. The upside of edge intelligence on the other hand is only Naturally, organisations that invest in edge intelligence just making itself apparent and many of the potential are doing so in the expectation of a return and the nature benefits are not yet widely understood. A system that can of that return is becoming clearer and clearer as the ingest, store and analyse data and then make decisions number of deployments increases. There have been that draw on machine learning and AI information while numerous trials and proofs of concepts targeting being located at the edge to ensure speed of response has augmented and virtual reality applications, local content huge value to add. It feels like today we’re on the edge of distribution, security and surveillance, manufacturing edge intelligence. Hopefully this issue of Tech Trends helps in pushing further over that edge. VanillaPlus Edge Intelligence I Vol. 2, No.3 I 05

6. Analyst report VanillaPlus Jim Morrish REPORT Transforma Insights EDGE COMPUTING WILL BE CRITICAL FOR DIGITAL TRANSFORMATION Technology enabled solutions are becoming ever more critical to the day-to-day operations of many enterprises. Among the most impactful technologies are artificial intelligence (AI), Internet of Things (IoT), cloud computing, and next generation communication technologies such as 5G. Edge computing may garner fewer headlines, but it is a key enabler for many solutions that utilise the emerging technologies listed above, writes Jim Morrish, a founding partner at Transforma Insights. SPONSORED REPORT VanillaPlus 06 I Edge Intelligence I Vol. 2, No.3

7. Analyst report Effectively, edge computing can allow local devices to operate to some extent autonomously of any cloud infrastructure Fundamentally, edge computing with certain devices able to call on • Regulatory compliance makes processing and storage processing resources residing in other Locally managed information resources available in close proximity nearby devices in a seamless way. potentially only needs to comply with to edge devices or sensors, local regulations, rather than multi- complementing centralised cloud • Improved robustness, resilience and jurisdiction regulations that might resources and allowing for analytics reliability apply in a cloud environment. close to those end devices. This results With more analytics undertaken in a number of benefits that can be locally to data sources, systems are not • Reduced operating costs very relevant in an enterprise context, as susceptible to disruption in the case Undertaking more analytics locally, including: that a connection to a remote cloud supported by edge computing, can location fails. Effectively, edge reduce the amount of data that needs • (Near) real-time responsiveness computing can allow local devices to to be sent to cloud locations for Analytics that may have previously operate to some extent autonomously processing, so reducing been undertaken in offsite cloud of any cloud infrastructure. In some communications costs associated with locations can potentially be supported situations, edge devices can operate data carriage. It also reduces the locally, avoiding the need for raw data almost completely autonomously and burden of processing that must be to be transferred to a cloud location independently of any connection to supported by cloud infrastructure and and for results of any analyses to cloud infrastructure. more importantly the amount of data follow the same path back to a local that needs to be stored in the cloud, device. Accordingly, the time taken for • Improved security and data reducing costs for cloud infrastructure. a system to respond to new input protection information can be reduced to near With more data processed locally, Edge locations, data lakes real-time. many security and privacy issues associated with transmitting data to and data streams • Improved device-to-device cloud locations can potentially be communications mitigated, and it can be easier for Thus far, we’ve referred to ‘the edge’ in Communications and the exchange of enterprises to demonstrate quite general terms as being data between devices that are co- compliance with data privacy and data characterised by deploying compute located together can be routed more sovereignty requirements. power closer to edge devices or directly, and without need to transit Alternatively, edge computing can be sensors. However, there are many cloud infrastructure. In fact, edge used to anonymise data locally before different kinds of edge location as intelligence can potentially allow onward transmission to cloud illustrated in Figure 1 overleaf. processing resources to be shared infrastructure. between a number of local devices, VanillaPlus Edge Intelligence I Vol. 2, No.3 I 07

8. Analyst report Device Edge Enterprise Network Data Centre Data Centre Edge Gateway Edge Edge Edge Core Data lakes Data streams Figure 1: Edge locations, data lakes and data streams The most local kind of edge versions of edge is another definition Typically, today most organisations computing is where compute power is of edge that has been adopted by will establish their analytic installed on an actual end device, for providers of cloud infrastructure and approaches and frameworks at example an industrial robot. Next which includes the provision of central, cloud locations. However, as most local would be an edge gateway, hosting capacity in secondary and described earlier in this report, there located close to an end device; often tertiary cities: closer to the end can be significant benefits to this would be an industrial computer devices, from the perspective of a undertaking these analyses closer to deployed to connect an operational cloud provider. We mention this for the end device. A specific example technology (OT) asset to information completeness only since, from the might be the case of AI-enabled technology (IT) systems. perspective of a solution designer, analyses of machine status such locations are to all intents and information, where powerful cloud The enterprise edge includes servers purposes still cloud locations. This is infrastructure can be used to establish and compute power that are local to entering a grey area though, since the rules to support, for example, pre- an enterprise, or on-site at an public cloud providers have been emptive maintenance alerts, and then industrial facility, and sit at the partnering with communications any rules identified can be deployed interface between that local network service providers (CSPs) to co-locate locally to significant benefit and and associated cloud infrastructure. the cloud edge and network edge updated dynamically. The first The network edge, specifically multi- together. dynamic associated with the advent of access edge computing (MEC), is an edge computing is a trend to push evolution of the enterprise edge In terms of the processing of data, data decisions closer to the edge. scenario, where edge processing is lakes and large databases of provided at the edge of a information tend to reside in the Conversely, raw information communications network. This cloud, whilst data streams can originating from a device can be scenario will be increasingly relevant potentially be processed at any of the redacted or filtered or aggregated as it with the advent of 5G edge locations described above. flows over edge infrastructure, communications, and particularly in ensuring that only necessary and the case of 5G private networks where Dynamics of data at the edge meaningful information remains for the network edge could be located in onward transmission. In the example the same place as the enterprise edge. The next aspect of edge to consider given above, once pre-emptive are the dynamics of information, the analytics have been deployed locally, Beyond these quintessentially local data, as it flows over edge assets. it simply isn’t necessary for the same raw data to be transmitted onwards to The most local kind of edge computing is where compute cloud infrastructure and the digest form could be sent instead to power is installed on an actual end device accelerate the central analytics that resides in the cloud. VanillaPlus 08 I Edge Intelligence I Vol. 2, No.3

9. Analyst report Device Edge Enterprise Network Data Centre Data Centre Edge Gateway Edge Edge Edge Core Analytics pushed to the edge Data lakes Redaction at the the edge Data augmentation Sweet spot: • Suitable compute power to support complex analytics • Low latency on local networks • Limited need to redact raw data • Advantageous location for data augmentation Figure 2: The dynamics of information in the context of edge assets, with sweet-spot Lastly, information flowing from edge end asset, but the information that • Data redaction devices can be augmented as it flows remains is generally richer and more Since local connectivity is essentially across various edge assets by meaningful. free within most facilities, there is associating contextual and other often limited benefit in redacting data information, often including These dynamics of information in the any nearer to end devices than the information from other local devices. context of edge assets are illustrated enterprise edge, while redacting at the For instance the industrial robot in Figure 2 above. enterprise edge can potentially save described above may have on-board significant costs associated with sensors to monitor operating processing data using cloud Enterprise edge as a sweet temperatures and computers located infrastructure. at the device edge or edge gateway spot level could be well placed to identify • Data augmentation when increasing device temperature Also highlighted in Figure 2, there is a The enterprise edge is an ideal is a result of some kind of sweet spot for edge computing, location to draw together all kinds of malfunction. However, relevant identified as the enterprise edge and data streams from within a local contextual information might be that which combines a number of benefits: facility, ranging from machine a factory air conditioning system has performance information to building failed, and thus it is likely that • Compute power information and more. It is also an increasing device temperatures are Significant compute power is readily ideal local location to draw in any simply a product of an increasing available at the enterprise edge, at a useful information from remote cloud ambient temperature. The potential reasonable cost. Compute power locations, including information like for this kind of data augmentation deployed at the enterprise edge is also weather forecasts or supply chain increases with distance from the end generally scalable, and additional status information. asset. For instance, enterprise edge is compute power and storage can be a good location from which to analyse added relatively easily. Overall, these factors make the and control an entire local facility, enterprise edge an ideal location for including buildings and building • Low latency undertaking all kinds of analyses of condition, the performance of any Analytics at the enterprise edge the wide range of information manufacturing lines, and the benefits from high-speed local generated by a local facility. There is organisation of work within the network connectivity within local also a particular synergy with the facility. facilities and also avoids the need to advent of 5G private networks, since communicate information to cloud 5G networks can support very low In general then, volumes of data tend locations. latency communications and MEC to reduce as they travel away from an can provide an efficient bridge VanillaPlus Edge Intelligence I Vol. 2, No.3 I 09

10. Analyst report Self-driving cars are a good example of edge computing being used to enable a local system to function autonomously without access to cloud infrastructure. between the private network and availability of parts, and the profile of The application of AI and machine compute power deployed at the orders that need to be fulfilled. learning (ML) to closed-circuit enterprise edge. Information can also be streamed to television (CCTV) feeds is a key augmented reality (AR) interfaces opportunity for edge computing, Tangible benefits of edge such as tablet computers, or video particularly given the proportion of goggles, with minimal latency, Such data communication volumes that it is The simplest applications of edge applications are currently a reality in a possible to redact by applying analytic technology are probably in building production manufacturing rules at the edge. AI enabled cameras control and facilities management, environment, but may also find can simply transmit an alert that a including heating ventilation and air application in smart cities, for certain decision rule has been conditioning (HVAC) control, and example, to control traffic, and also triggered, averting the need to security monitoring, access control healthcare environments, for example, transmit a full moving image feed to and alarm systems. In a post-Covid to manage patient flows. Streaming cloud infrastructure for analysis. Such world, such solutions could also information to an AR interface can be solutions are commonly productised extend to include air quality and particularly useful in construction and in the form of facial recognition ventilation monitoring. Clearly such field support contexts. cameras, or security cameras, and scenarios benefit not only from autonomous operation at the edge, but also the location of analytics at the enterprise edge allows for good oversight of all parameters associated with a particular facility. This kind of application is clearly applicable in almost any vertical sector. Another potential application is production monitoring and control in an industrial context. For example, an edge-enabled system can be used to monitor a range of production lines and machinery in a manufacturing location to pre-emptively identify any maintenance that is required, and dynamically re-scheduling production given anticipated downtime, VanillaPlus 10 I Edge Intelligence I Vol. 2, No.3

11. Analyst report “ The simplest applications of edge technology are probably in building control and facilities management... ” clearly have application in almost all significant relevance beyond the vertical sectors. utilities sector including, for instance, agricultural and Self-driving cars are a good example aquacultural applications, of edge computing being used to transportation management - enable a local system to function particularly rail and public autonomously without access to cloud transport, oil and gas infrastructure. The same concepts are extraction and relevant in other semi-autonomous mining operations. vehicles, ranging from factory floor robots to drones, and from automated transportation systems to hospital Conclusions machinery. Many of the applications associated with these kinds of devices In conclusion, many variants of edge are in fact more complex than computing are going to be critical to autonomous cars, since they must industrial and commercial operations combine local, autonomous, in the near future and edge will be a operations with coordination with key technology for enabling Industry control applications hosted on cloud 4.0 and many similar developments. infrastructure. Within the edge space, it is the enterprise edge that will be the focus Another application that is of most attention, even more so in the particularly suitable for edge context of 5G private networks. computing is the management of an estate of devices, for instance as might be found at a solar farm or a wind farm. Local edge computing devices would be able to compare the performance of individual solar panels or wind turbines, identifying those Report sponsor that need maintenance and ensuring appropriate performance of the overall device estate. Again, such an application concept could have VanillaPlus Edge Intelligence I Vol. 2, No.3 I 11

12. Sponsored interview Multi-access edge computing (MEC) is coming together with the low latency mmWave connectivity enabled by 5G to make intelligence at the edge a reality that truly enables bi-directional digital conversations and gives the potential for technologies such as digital twins to be operated effectively. Here, Dheeraj Remella, the chief product officer of VoltDB, tells George Malim, the managing editor of VanillaPlus, how the challenges associated with MEC are being addressed and why, for many, once they’ve looked over the edge, there will be no going back to centralised, remote architectures. 5G AND MEC ENCOURAGE EDGE INTELLIGENCE TO BECOME PERVASIVE VanillaPlus 12 I Edge Intelligence I Vol. 2, No.3

13. Sponsored interview Dheeraj Remella VoltDB George Malim: Does the latency offered by response instructions flow in the digital-to- mmWave make the goal of linking physical physical direction. This bidirectional digital and digital twins in real-time a reality? What conversation completes the picture of the are the complex impacts of this on real-time control loop that is going to be a enterprises and can data really be utilised fundamental requirement for successful live in the stream of enterprise processes? transformation based on machine-to- machine communication. Dheeraj Remella: There are a few things that are coming together in this scenario. First, The factors in this scenario profoundly there is the high bandwidth provided by 5G impact how enterprises see the value of data. deployments. 5G has three bands of Data’s value for analytics is well established operation: low band, mid-band and high by now. But what is hidden is the value of band. High band/mmWave allows massive data captured in the first few milliseconds. amounts of data to be transferred over mobile The low latency event-driven servicing networks. But the caveat is that mmWave becomes especially significant when we are doesn’t travel too far and is easily talking about the digital automation of interruptible by obstacles. So, an ideal business processes. Just a few examples of environment for mmWave to be used the hidden value from our experience at optimally would be within a building. VoltDB are an 83% reduction in fraudulent transactions completing, 100% prevention of This caveat brings us to the second element distributed denial of service (DDoS) attacks of the scenario, which is digital on the network, and 100% detection of bots transformation. Enterprises often embark on before they intrude into a bidding process. All a digital transformation journey but end up of these have stringent latency service level trapping themselves into what I call digital agreements (SLAs) to make these decisions. transliteration. Digital transliteration is when Data usage is going to augment post-event an enterprise simply digitises its processes analytics by using that intelligence for in- status quo using modern technology. I see event decision making. Low latency this as a wasted opportunity to optimise decisions impacting operations require low business processes to break free of latency availability of event data. mmWave traditional constraints and evolve to operate in private 5G settings will accelerate the shift in a realm of new possibilities. to tapping into the first ten milliseconds of data’s life. The third element here is the need for closed- loop digital twins. Historically, digital twins GM: What are the challenges of extracting all were state-recording systems whose data the value from the platform? was heavily used in analytics to understand the systems, assets, and process behaviour. DR: Most enterprises have already invested But given the drive towards automation, in many technologies to fulfil various forms digital twins are now the intelligence behind of value extraction from data. When faced their physical counterparts. Data is not just with the challenge of tapping into the real- flowing one way from the physical asset, be it time data streams, these enterprises often equipment or people or even a business resort to figuring out how to make-do with process, to its digital representation. The flow these existing investments. But they fall into needs to be bidirectional, where data comes the trap of spending inordinate amounts of from the physical-to-digital path, and the time, money, and resources, only to end up VanillaPlus Edge Intelligence I Vol. 2, No.3 I 13

14. Sponsored interview making compromises, or even outright event took place, the event data needs to customer premise equipment. There are failing at the attempt. The low latency travel quite a bit before realising its value. steps being taken to bring MEC to the expectations for event-driven real-time network edge, and we’ve already value extraction from data require many The real-time needs of digital observed that major communications different capabilities to play together in a transformation are pushing the edge service providers (CSPs) are partnering single unified technology. intelligence agenda forward in an with public cloud vendors to get the cloud accelerated manner. Now the question is, experience to the CSPs’ data centres. In When patching together various where is edge? You have the edge in either case, MEC is on single unit boxes technologies, enterprises face pitfalls like: devices, gateways, customer premise and not taking advantage of all the equipment (CPE), network edge, and then distributed computing architecture - Communication latency between there is the central/cloud data centre. innovations. When MEC becomes an various layers breaking SLAs MEC is going through evolution itself. It is integral part of an enterprise’s operations, - Infrastructure footprints bloating transforming from a simple aggregator or MEC needs to provide the ability to scale because each layer requires its local data storage to become more and for business continuity by clustering resiliency model actively involved in edge intelligence. In for resilience and performance. The - Complex failure handling when my opinion, MECs are going to get a lot initial thought that comes to mind is ‘why aiming to maintain business more capacity allocated to accommodate invest in near-edge infrastructure when continuity increasing responsibilities. It might even one can have all the hardware necessary - Operations using stale machine change the definition of MEC. procured and managed centrally, learned insights especially on top of the central data - Outdated decisions because of not MEC on CPE will be the perfect place to centre’s massive investment?’. Right after meeting latency SLAs land a variety of capabilities: that comes the concern of security and - Data thinning intrusion prevention. On the other hand, a unified platform - Automated event-driven decision ingests, stores, analyses and makes making My thoughts on this are: if MEC and decisions, which are continuously - Data preparation for analytics - This consequently edge intelligence is enriched with machine learning becomes even more significant when you implemented well for the right use cases, retraining cycles, to address all needs consider reducing the infrastructure the returns will easily outweigh the cost. holistically. costs at the central data centre when not Besides, the intelligence at the edge can all raw data needs to be stored even potentially decrease the GM: Even with the low latency, is the - Incorporation of machine learning infrastructure needs in the central cloud still too far away to enable the insights into the decision-making datacentre. As for security, it is not round trip to be completed in time to process something that is an afterthought. Safety create a meaningful interaction? Does is an integral part of everyday operations this mean MEC is a must? A plan to succeed in bringing intelligence and weaves into business operations. A to the edge must include increasing the no-trust network implementation DR: This question is an exciting one for role of MEC. intertwines checks and prevention into me. Cloud computing brings a lot of value every step and interaction. Now, the to commoditising compute, storage and GM: What are the challenges to MEC that biggest challenge is the willingness, or networking. But, these new low latency need to be addressed? Are costs, security lack thereof, of enterprises to undertake expectations and value forces the and the definition of the edge itself still this journey and not shackle themselves intelligence closer to the edge. Cloud barriers? with old ways because that is how it is vendors, for a large part, are centralised today. in large data centres. This centralisation DR: Currently, MEC is underutilised, means that, immaterial of where the although it is in the right place, in the VanillaPlus 14 I Edge Intelligence I Vol. 2, No.3

15. Sponsored interview “ The real-time needs of digital transformation are pushing the edge intelligence agenda forward in an accelerated manner ” GM: Is moving intelligence closer to the What is the shortest path from an event edge the new standard? Is there no going to a responding action? back? What is necessary to drive those actions DR: The bottom line is there is no denying intelligently? that enterprises are always looking for an edge to take their business to the next How can we do this while using the least level and differentiate themselves from amount of hardware and resources? their competitors. With Industry 4.0 getting accelerated by 5G, traditionally I would highly encourage our audience to non-digital companies are looking to check out our paper on Intelligence at the optimise their business processes and Edge. utilise communication and computing technology advancements. Innovative www.voltdb.com leaders that adapt to this fact will leave enterprises that do not behind. Edge will become the most explored area of innovation, bringing better security, optimisations, and user experience. Once the path to the edge is taken, there is no going back. GM: So where does VoltDB play in all of this? DR: VoltDB has meaningfully integrated in-memory database and stream processing technologies. This combination brings the best of both worlds, such as data consistency and fast storage with transaction processing from the database world and the ability to integrate with streams for ingestion of data and communication with other applications and systems. Our engineering team has built our technology as a single cohesive product instead of just assembling various open source technologies and calling it a platform. Every step of our design process considers three inextricably linked questions: VanillaPlus Edge Intelligence I Vol. 2, No.3 I 15

16. Low latency decisions WHY LOW LATENCY DECISIONS ARE KEY FOR EDGE SUCCESS 5G, edge computing, digital transformation, digital twins, machine learning and AI appearing together looks like a buzzword bingo. But there is a rational connection between these topics and technologies, writes Dheeraj Remella, the chief product officer at VoltDB. To frame these topics in simple terms, they can be defined full sense control feedback loop as follows: Machine learning - Employing computers and algorithms 5G - Provides high bandwidth low latency connectivity to to understand what the data is telling us the mobile network, which can be private or public depending on the use case Artificial intelligence - Operationalising the machine learning insights to Edge computing - Brings intelligence meaningfully closer facilitate ever-evolving ‘do- to the source of events learn-do better’ cycles Digital transformation - An exercise that an enterprise undertakes to break free of archaic processes and redefine operations to make use of modern technology Digital twins - The next step in the digital representation of organisational assets, such as people, assets and processes, moving from simple state recording to SPONSORED ARTICLE VanillaPlus 16 I Edge Intelligence I Vol. 2, No.3

17. Low latency decisions Now, with this backdrop of understanding of the terms, let’s from the network edge. But in either case, examine the fundamental thread that connects these while the network ping roundtrip latency is together. The singular intent of each one of these important, the service latency in the technologies is to bring more self-learning automation to middle of the communication business processes. Data generated by enterprise assets and roundtrip is also essential. processes should get consumed as close to the source of the events as possible. This ensures decisions are not made - and actions are not taken - on stale information. This would render those decisions and actions obsolete as the universe has since moved on. The decisions are not made based on a static set of rules. Instead, they have a characteristic of dynamism due to the speed at which events get generated and decisions are made. The overarching business To keep the service latency low, a data platform that serves multiple purposes together is necessary to ensure the data value is extracted without an artificially complicated technology stack. The event data needs to be ingested, stored and aggregated, either as a single business event or a set of events - think complex events, and compared to the aggregated data to measure some key performance metrics. Any deviation in behaviour needs to be acted upon for either monetisation or prevention of some form of threat. process is Monetisation opportunities can range from the fluid to adapt personalisation of user experience to determining the most to current profitable end product - at this moment - from refined crude conditions. These oil. Threats can range from potential machine downtime to dynamic rules are robotic network intrusion. Given the core objective of digital generated by training and transformation is to automate business processes by retraining learning models. shifting to a machine-to-machine communication paradigm, the latency of these decisions to invoke When dealing with real-life processes and appropriate actions needs to be in single-digit milliseconds industrial applications, ‘good enough’ is not and, in a more stringent environment, less than a enough. Every decision and action needs to be millisecond. If this window is missed, a cascade of correct based on the current situation and business inefficiencies is put into play. The decisions are stale, which rules. This is where bringing intelligence beyond simple translates into wrong learnings. This in turn creates insight aggregation and to the edge becomes important. Now, what that is not congruent with the reality of the enterprise. is edge though? Is it in the device or gateway or some kind of on-premise datacentre or network, or is it in the cloud? It Depending on the use case, VoltDB’s customers typically depends on the application. But in most cases, one thing allocate anywhere between 0.25 milliseconds to less than 10 becomes concretely evident. The edge on the devices is too milliseconds for making the decisions, and those decisions narrow to make any meaningful contextual decisions. are made with 100% accuracy and uncompromising guarantees for resiliency. Completing intelligent decisions Having the edge layer in the cloud is too far away to be able and taking action within this timeframe is no longer a nice- to act on events within a reasonable amount of time. to-have. It is a must-have to uncover the latent value of the Gateways are slowly becoming irrelevant with narrowband data in its infancy. Our customers are able to gain IoT and Cat-M devices being able to connect directly to the unprecedented advantages ranging from being able to network through embedded (eSIM). So that leaves the on- prevent 100% of bot intrusion to being able to take the best premises data centre and the network edge as the best next action by adhering to stringent low latency service possible candidates for intelligent interactions. level agreements (SLAs). Applications that are more industrial by nature will best benefit by taking the on-premises approach while www.voltdb.com applications that are more consumer-oriented will benefit VanillaPlus Edge Intelligence I Vol. 2, No.3 I 17

18. Securing the edge CAN IT BE SAFER ON THE EDGE? Children are warned to stay away from the edge of cliffs, bridges, roads and rivers, but in computing, the edge is turning out to be one of the safer locations for hosting processing power and analysing data. Antony Savvas explores why this is the case, while also addressing the security concerns that do exist around edge computing and assessing how these are being resolved. The edge computing market includes hardware, application performance by bringing processing edge nodes or gateways and servers, sensors and tasks and running applications closer to the routers, software including databases and cellular customer. The implementation of MEC at analytics, services and edge-managed platforms. mobile base stations or edge nodes is expected to The global edge computing market size covering facilitate the rapid and flexible deployment of new all these areas is anticipated to reach US$43.4bn services and applications for customers, which by 2027, seeing a CAGR of 37% over the forecast "promises healthy market growth", Grand View said. period, according to a March 2020 report by Grand View Research. Furthermore, there is an anticipated wave of micro edge data centre (EDC) capacity that differs from Covering just hardware, platforms and services, large centralised data centres. This new capacity rival research house MarketsandMarkets projects is expected to range from small clusters of edge the global edge computing market will grow from cloud resources located on street-lights to a few US$3.6bn in 2020 to US$15.7bn by 2025, at a CAGR racks located in a shelter at the base of a cell tower of 34% during the forecast period. or inside buildings. Some key players in the edge computing market are ADLINK, Altran, Amazon Web Services (AWS), Axellio, Belden, Cisco Systems, Clearblade, Dell Technologies, Digi International, EdgeConneX, Edge Intelligence, Edgeworx, FogHorn Systems, GE Digital, Google, Hewlett Packard Enterprise, IBM, Intel, Juniper Networks, Litmus Automation, MachineShop, Microsoft, Moxa, Nokia, Sierra Wireless, SixSq, Vapor IO, VMware and VoltDB, among many others. The main drivers 5G is expected to act as a big catalyst for market growth. Applications using 5G are expected to change traffic demand patterns, "enabling technology growth avenues" for the communications service providers (CSPs), said Grand View. The cloud market leaders see this as a threat and have started investing in the edge ecosystem themselves by engaging in partnerships with CSPs. CSPs are expected to embrace new opportunities in the multi-access edge computing (MEC) market place, said Grand View. MEC allows providers to mitigate network congestion and ensure higher Antony Savvas VanillaPlus 18 I Edge Intelligence I Vol. 2, No.3

19. Securing the edge In addition, 5G networks can use EDC bringing the intelligence even closer to ongoing pandemic has fostered the need facilities to provide efficient local data their premises to support lower latency for business continuity plans that services, redirecting edge traffic away applications and deliver improved include flexible, anywhere, anytime, from the carrier networks to local public security." secure remote access at scale, even from internet networks. Various start-ups, untrusted devices," said Gartner analyst such as EdgeMicro, are in the process of So is it just 5G and the Internet of Things Joe Skorupa in August 2020. "Mobile deploying commercial mini data centres (IoT) that have driven the edge data workforce, contractor access and edge with IT computing stacks, redundant processing and analytics market? Or has computing applications that are latency cooling, fire suspension and biometric this been an evolving migration driven sensitive are three market opportunities. security. by the needs of communication service Over the last three months, SASE has providers and enterprises, and if so, what been adopted by more than 40% of global MarketsandMarkets warns however that else do they want out of the edge? remote workers." the costs of moving to the edge can be significant. It says: "Edge computing "5G and IoT are accelerants to this Many suppliers have already launched might reduce data transmission and awakening to the real-time needs of SASE-based products onto the market, storage costs through localised digital transformation,” Remella explains. with VMware one of the latest to do so in processing, but investing in edge “Customers already in the space of September 2020. The VMware SASE infrastructure still adds to the capex of tapping into event-driven real-time Platform converges cloud networking, companies, including heavy investment decisions and automation have benefited cloud security and zero trust network in edge nodes, other edge devices and from revenue increases and greater access with web security, to deliver edge data centres." security. These organisations were flexibility, agility and scalability for pioneers because they saw the value of enterprises of all sizes, said the vendor. They would also be required to spend what is being ignored in the first ten With the edge data processing/analytics more on making the devices and the milliseconds or less.” market expanding as billions more entire network secure. But, things are added to the edge, we are MarketsandMarkets added: "The edge "What 5G brings to the table, especially inevitably going see more edge data infrastructure cost is a restraining factor, when combined with narrowband IoT leakages and security mishaps hitting though, with advancement and (NB-IoT) and Cat-M, is the ability to do the headlines. continuous R&D, the cost of edge away with hops and interim aggregators technology is expected to reduce soon." to get to the network directly,” he says. To mitigate matters VoltDB's Remella So despite some obstacles, there is “Gateway-less IoT is going to become says edge intelligence is going to be a plenty of interest in processing data at mainstream and this will allow much central theme in data-driven enterprises. the edge. But why is it safer, overall, for richer intelligence near the edge, be it He says real-time dashboards at receiving, processing and exchanging machine learning or event-driven real- operations centres are going to give way data - instead of relying on cloud data time decisions." to automated processes that will centres to do everything? decrease the burden of manual It’s time to get sassy intervention and resolution. Edge safety and With more data moving to the edge, last "Machine learning is going to slowly performance December analyst house Gartner start moving elements from the central promoted the Secure Access Service data centre to the edge data centre,” he Dheeraj Remella, the chief product Edge (SASE) framework. Pronounced explains. “What if we can do away with officer at edge database and analytics sassy, SASE enables increasingly all the unnecessary raw data and instead provider VoltDB, says: "It is pure and distributed and mobile workforces to digest it locally near the edge? And the simple. Getting closer to the event source remotely and securely access corporate central data centres are intelligence decreases the time elapsed before the networks and clouds. Interest in SASE aggregation points where only the event data becomes stale. You can apply has grown substantially due to Covid-19 learnings from the edges need to be this to a variety of value extraction as enterprises recognise its potential as a mixed together to create higher level principles such as personalisation, business continuity solution. models, which can then be sent back to operational automation, preventative individual edge centres to incorporate maintenance and most importantly SASE combines network security into local decisions?" securing assets, processes and even functions - such as secure web gateway employees and customers." (SWG), cloud access security broker This would reduce the amount of data (CASB), firewall as-a-service (FWaaS) and being stored in cloud data centres, "While the central cloud data centres zero trust network access (ZTNA) - with reducing energy consumption, with offer larger infrastructure capacity, just software-defined wide area networks organisations only having to store the travel time for the data to get to the (SD-WAN) to support the dynamic secure human transactional data covering data centre robs the enterprise of the access needs of organisations. These financial or medical or insurance opportunity to respond to the event in a capabilities, says Gartner, are delivered matters, for instance. "I still encounter timely manner,” he adds. “You can primarily as-a-service with the ability to enterprises talking about storing several observe this emphasis on getting close to identify sensitive data or malware and petabytes of data and the need for the event source in efforts by CSPs the capability to decrypt content at line massive high-performance compute partnering with cloud vendors for the speed. clusters to churn through that data,” says deployment of edge data centres. And Remella. “People might call me utopian, enterprise customers stand to benefit by "Although SASE is relatively new, the but I strongly feel we can get there." VanillaPlus Edge Intelligence I Vol. 2, No.3 I 19

20. Edge analysis PANDEMIC PUTS PRESSURE ON ORGANISATIONS TO PUSH INTELLIGENCE TO THE EDGE Multi-access edge computing (MEC) has turned the corner from being seen as an interesting alternative to the hub and spoke architecture of cloud computing to being recognised as a vital technological enabler of low latency, connected intelligence. George Malim examines the latest analyst predictions for the technology’s increased uptake. It comes as no surprise given the drivers that are immediate growth. It continues to expect the affecting the technology market that substantial market to start slowly before accelerating growth in MEC is expected. The arrival of 5G, with significantly in the second half of the 2019-2024 its low latency communications capability, the forecast period. The firm expects China to be the dispersal of more and more people and endpoints to leading market in terms of the scale of MEC the edges of networks, and the increased familiarity deployments and the country will represent the with applying intelligence at the edge have all come largest regional market. together to provide important stimulae for growth. This is borne out by research from Dell’Oro Group Frost & Sullivan’s recent analysis, 5G and Edge which projected in its July 2020 forecast that the Computing—Cloud Workloads Shifting to the Edge, MEC market will grow at a compound annual Forecast to 2024, also finds that edge computing is a growth rate (CAGR) of 169% between 2019 and 2024. foundational technology for industrial enterprises because of the lower latencies, robust security, Coronavirus-related lockdowns and new ways of responsive data collection and reduced costs it can working are demonstrating that more needs to be offer. The firm reports that, in spite of being at a done from remote locations and, with office work nascent stage, the MEC market is estimated to grow now moving away from large, centralised locations, at a CAGR of 157.4%, garnering a revenue of so too are computing resources. Dell’Oro’s latest US$7.23bn by 2024, up from US$64.1m in 2019. estimate is a revision of its January 2020 forecast and reflects the industry collaboration and “The recent launch of the 5G technology coupled momentum being achieved. The firm says that this with MEC brings computing power close to traction has caused it to double its predicted CAGR. customers and also allows the emergence of new Nevertheless, the firm does not project steep applications and experiences for them,” said “ ...many believe lockdowns and quarantines have only accelerated digital transformation in general 20 I Edge Intelligence I Vol. 2, No.3 ” VanillaPlus

21. Edge analysis “5G and MEC are an opportunity for telecoms operators to launch innovative offerings Renato Pasquini, the Information & Communication Technologies research In fact, the importance of edge is being recognised to the extent that ABI ” chipset market in 2025.” director at Frost & Sullivan. “5G and MEC Research expects AI chipset revenues The pandemic has disrupted demand for are an opportunity for telecoms from edge deployments will dethrone many smart consumer devices, notably operators to launch innovative offerings cloud as the leading market by 2025. At smartphones, smart home, and and also enable an ecosystem to flourish that point edge AI chipsets will generate wearables. These would have helped in the business-to-business (B2B) revenues of US$12bn per year, outpacing further stimulate deployment of AI segment of telecoms service providers the cloud AI chipset market, which will accelerating technologies at the edge using the platform.” reach US$11.9 billion in 2025. and, at the same time, implementation of AI in industrial manufacturing, retail and A recent Heavy Reading survey has also Cloud remains at the centre of AI today other verticals has been postponed or uncovered that communications service with most workloads served in public put on hold. This is likely to be only a providers (CSPs) see MEC as an and private clouds. However, the temporary state of affairs. opportunity. Almost 85% of network industry is beginning to shift away from operator professionals polled believed centralised cloud resources driven by the “ABI Research expects the market to that edge computing would be critical or need for privacy, cybersecurity and low rebound in 2022. It is important to note important to their network evolution latency and AI training and inference that the impact on the chipset supply strategy, the firm says. workloads on gateways, devices and chain has been relatively minimal since sensors is starting to happen. Recent fabrication factories in Singapore and For Pasquini, the CSP opportunity comes advancements in key domains, Taiwan remained operational during the second to the edge applications and including connectivity to cloud outbreak,” Su points out, adding that software market. “From the perspective computing, new AI learning architecture vendors of key connectivity technologies of the MEC ecosystem, software - edge and high-performance computational such as 5G, Wi-Fi 6, and autonomous applications and solutions - promises chipsets have played a critical role in solutions such as autonomous vehicles the highest CAGR followed by services,” this shift, the firm says. see minimal impact to their product he said. “[These are composed of] roadmaps. “Catalysing many other telecoms operators’ services, cloud “As enterprises start to look for AI emerging technologies, edge AI will pave providers’ infrastructure-as-a-service, solutions in the areas of image and the way for a variety of new business and edge data centre colocation object recognition, autonomous material opportunities in the consumer and services.” handling, predictive maintenance and enterprise segments.” human-machine interface for end MEC is not simply a CSP opportunity. devices, they need to resolve concerns Covid-19 may have caused a slight pause Frost & Sullivan predicts that around data privacy, power efficiency, in production of technologies that pave approximately 90% of industrial low latency and strong on-device the way towards intelligence at the edge enterprises will utilise edge computing computing performance,” explains Lian but it has also made people look in more by 2022, presenting immense growth Jye Su, a principal analyst at ABI detail at the edge as they assess what prospects for MEC market participants. Research. “Edge AI will be the answer to the new world of work will look like. In Consulting firm, Deloitte, has reported this. By integrating an AI chipset addition, many believe lockdowns and that more than 80% of executives designed to perform high-speed quarantines have only accelerated surveyed in a recent report believe that inference and quantized federated digital transformation in general and advanced connectivity is very or learning or collaborative learning that can only further accelerate the extremely important to their ability to models, edge AI brings task automation march towards greater intelligence at capitalise on advanced technologies and augmentation to device and sensor the multi-access edge. such as artificial intelligence (AI), edge levels across various sectors. So much computing and data analytics. that it will grow and surpass the cloud AI VanillaPlus Edge Intelligence I Vol. 2, No.3 I 21

22. Edge ROI MEC PROVES THE NEED FOR SPEED There are some very sensible, industrial strength cost justifications for installing multi-access edge computing (MEC) but sometimes attention can be diverted by more fanciful projections, writes Nick Booth. Forget about augmenting and virtualising realities, emerging practical use cases in manufacturing. there is plenty of work to do in the factories, Sensitive industrial process-control systems will according to Dean Bubley, founder of Disruptive ruin their finely-calibrated machinery if they wait Analysis. “A lot of what gets discussed in 5G and ten milliseconds to respond. Image sensors and edge-computing conferences is either hyped or network sync mechanisms work in nanoseconds undeliverable,” he says. “Many of the use-cases can and a picosecond is a long time to a Photon sensor. be adequately serviced with 4G mobile or Wi-Fi - or Ultra-fast laser pulses for machining glass or a person on a bicycle delivering a USB memory polymers need feedback in femtoseconds. stick.” Kiva Allgood, the head of IoT at Ericsson, agrees However, teleprotection systems for high-voltage that manufacturing is the most realistic target utility grids will demand latency of between market for MEC, and identifies fabrication and 3D 6-10ms. Bubley has identified a range of other printing as avid consumers. VanillaPlus 22 I Edge Intelligence I Vol. 2, No.3

23. Edge ROI Meanwhile, software vendor VoltDB reports areas Can a 5G network deliver the same in which clients face problems now. “It’s hard to bill accurately under the present circumstances benefit as MEC? and a local presence would help,” says Dheeraj Eventually, yes, says Ericsson’s Allgood. In theory Remella, the company’s chief product officer. Releases 15 and 16 of 5G will deliver the low latency and high frequency connectivity needed Security, too, is becoming unfeasible without a for IoT. rapid response. Nobody can defend against a distributed denial of service (DDoS) attack if their Anyone who waits for 5G will be disappointed intelligence has to be sent to a distant data centre says Joshua Norrid, senior technical director at to be processed. All reporting and reconciliation of DataStax, because many 5G standards don’t intrusions must be instant. Similarly, automated specify how data is stored, managed or replicated. processes that are fine-tuned by artificial intelligence (AI) will falter if the machine learning “Enterprises want a consistent approach to Joshua Norrid managing data from the edge to centre and in- DataStax is delayed, says Remella. between and they won’t find this model flexible enough,” he says. “If it’s hard to use your data, then Communications service providers (CSPs) can’t developers will struggle to build the IoT work out hyper-personalised offers to each applications that will need it.” subscriber and capitalise on the mass marketing campaigns unless they can live in the moment. Dr Paul Carter, the chief executive of mobile Which calls for some heavy lifting at the edge. network benchmarker Global Wireless Solutions, urges caution over your choice of MEC. “Think There are exceptions. No credit card company can about the connectivity charges: is it cheaper to run a full fraud inference routine on a server at the connect via Wi-Fi or a non-cellular network,” says edge because there is too much information and Carter. context to process. In this case, not every event is significant, so the client can thin the data out Still, smartphones and tablets will eventually before sending it to a central data centre. bring AR apps to the masses at which point CSPs Dr Paul Carter will have to compete for subscribers on their No company should digitise any process if they Global Wireless Solutions ability to match the digital picture to the physical, can’t guarantee its safety, says Remella, and MEC predicts Kevin Riley, the CTO at Ribbon can be a reassuring presence. Security probes and Communications. Another increasingly popular the cross-checking of events are now inextricable security app, face recognition for site access, will parts of the new process definition. The meta-data need MEC as it processes data locally. The describing these jobs can triple the amount of systems will fall down if they need to send event information generated. buckets of raw data back to HQ. The cases for MEC needs clarity and that is Neither will be realistic unless the computer provided by definition, says Remella. Edge is a comes back with an answer in under 12 contentious term and it’s possible to get lost in the milliseconds, and “Ideally less than seven,” semantics of defining in-devices, gateways, says Riley. customer premises and network edges. It’s better to ignore these philosophical topics – the question Latency is important, for application developers, that matters is: How close can I get to the event enterprises and many classes of IoT device and before it peters out? solution. But we have been spectacularly vague at defining what low-latency actually means, and Thomas Neren, Ericsson’s head of dedicated where it's needed, says Disruptive Analyst Bubley. networks, is another who decries the lack of focus What is likely is that average latencies will fall and warns that people are getting confused over with 5G. An app developer that currently expects a 5G’s potential to meet the challenge. 30 to 70ms latency on 4G, and probably lower on Wi-Fi, will gradually adapt to 20-40ms on mostly- “We need to standardise our definitions and make 5G networks and eventually 10-30ms. If it's a sure we are all talking about the same thing,” says smartphone app, they likely won't use ultra- Neren. “In 5G terms, one millisecond is the latency reliable low latency comms (URLLC) anyway. of the radio interface only, whereas a measure of Specialised IoT developers in industrial settings total, end-to-end latency is the sum of response will work with specialist providers, like mobile times across the transport layer, the core, cloud network operators, fully-private networks and storage and computing systems.” integrators, to hit more challenging targets. “Returns on investment and safety constraints Creating ultra-low latency across a high capacity will justify the cost,” says Bubley. They may even wide area, would take a massive investment and get down to one millisecond at some point in the the natural starting point for confined areas might medium term, but it's far from clear they will be be dedicated networks, says Neren. contributing massively to the edge-providers' bottom line. VanillaPlus Edge Intelligence I Vol. 2, No.3 I 23

24.NE W eB O OK THE HIDDEN INFLE L CTION POINT O IN 5GG WIll the chang ging definition of reaal-time break your exis i ting tech stack ? Dive into our in-depth analysis, including 5 points to heelp you evaluate yourr 5G readiness. GRAB YOUR Y COPY OLTD T B.COM

0 点赞
0 收藏
0下载