Written by Simon Kenny, Chief Strategy at Hoptroff
Digital transformation and automation are changing the way all industries work, but with recent big increases in the processing power of microchips, falling costs of chip production and rapid acceleration of AI, machine learning (ML) and cloud-based IoT applications, the pace of change is set to speed up.
Over the next couple of years, we will see a decentralisation of data processing away from cloud centres and toward multiple edge computing centres. From there, AI and machine learning applications can deal with customer interactions independently, sharing only small volumes of data back to the central cloud, so that the explosion in new data coming from the Internet of Things (IoT) devices across 5G networks does not overload telecoms and data storage capacities. In the future of smart cities communities can improve energy distribution, streamline rubbish collection, decrease traffic congestion, and improve air quality with help from the IoT.
Every location and mobile and IoT device must share the same accurate timescale as the data centres they report into if data quality and AI and ML efficiency are to be optimised for different industries in the edge. If they don’t, then the time sequences in the data records will be incoherent and unreliable for applications to use. To prevent this you need standardised timescales, but the standards needed and the benefits to be derived will vary across different industries. Take IoT, 5G, autonomous cars, metaverse and intelligent network of connected objects and machines, for example. Each of these industries will be major contributors to the automated, smart city framework, but their time synchronisation needs will all be different.
Accurate and precise time synchronisation across 5G networks
5G telecoms needs an accuracy level of 1µs (one microsecond) to be maintained across the network to manage bandwidth and ensure different data streams on the network remain separate from one another. This is not a choice; it is a technical necessity to keep the network running efficiently. Without it, calls would be lost, and data streams fail. To achieve this very high standard the network requires GPS synchronisation and local clock hardware be deployed and maintained across the network.
5G networks and IoT are crucial to realise the smart city vision. They provide dependable and cyber secure hardware and connectivity to make these systems of connected objects and machines function. The benefit to 5G and its customers of investing in the infrastructure to maintain such a strict standard is that 5G can manage a very broad spectrum of bandwidth that gives it high reliability, low latency and an ability to support new areas such as industrial automation. This is something that 4G could not serve in the past. For businesses and consumers to reap the rewards of 5G, and for smart city operators to make the most from the 5G networks, investing in accurate and precise time synchronisation and the ROI associated with this is indisputable.
Getting the most from IoT device data
In IoT, the importance of accurate time is not so accepted. If connected traffic lights, autonomous cars and smart buildings do not all share the same time or cannot prove their time is right, that does not stop them performing their function. It does, however, undermine the quality and utility of the data they provide about their own function and how their data can integrate with data from other devices.
If you needed to prove a sequence of events from IoT data records of different devices was true, such as a door was opened at a certain time using an ID card, then later that the same person was captured on remote CCTV camera footage and then still later, that the same person was recorded speeding by a smart traffic cam in another location, you need the different devices to all share the same timescale.
If they don’t, the timestamps might tell you the person driving the car could not have possibly opened the door because the timestamp on the traffic cam says they were miles away at exactly the same time. The difference between the clocks undermines your ability to prove a sequence of events. However, if all the devices are united in a common timescale, verified by a Primary Time source, you can use all your IoT data as a reference point without any need for further verification.
Accurate time might not be essential to all IoT functions, but it is essential if you want to make full use of the data IoT generates. A standard of 10 milliseconds would cover the functionality of most IoT devices, but you do need a regular system of comparisons to a trusted time source, so you can verify the time is right and keep records so you can prove it later if required.
The rise of the smart city is demanding accurate and precise time synchronisation
Smart cities need accurate time because it’s both essential to their operation and because they need to be able to track and report on events.
To enable autonomous vehicles including automated, driverless and robocars to operate safely within the smart city, accurate time to a standard as high as telecoms, one microsecond, is essential. The internal components of the autonomous vehicles need to share the same time, the autonomous carsmust be synchronised with connected sensors in its external environment, and it must be synchronised with other autonomous cars, smart streetlights, and connected traffic cameras.
The Law Commissions of England and Wales and Scotland have recently recommended that manufacturers should be made liable for accidents that happen when the self-driving capabilities are in control. If this recommendation is turned into law, it is vital that autonomous vehicles keep records with verified timestamps of its automated actions, so that, in case of an accident, insurers can assess what happened and assign responsibility accordingly.
Accurate and precise time synchronisation is a necessity for increased automation
As automation takes on greater responsibility for dealing directly with the public, either as 5G, a smart IoT device or a component of a smart city like an autonomous vehicle, smart streetlights, or connected traffic cameras, it will also need to accept greater accountability. Some automated functions may not need accurate time to continue to work effectively, but all automation needs accurate time to ensure that it has data records of sufficient quality to explain/justify machine actions after the event.
The accuracy and reporting standards will not need to be the same for every activity, internet delivered time will cover many needs, but industries that want to extract maximum benefit from automation and build maximum trust with their customers, will adopt enterprise-wide smart and precise time synchronisation without the need for any regulator imposing it. Smart timing will just be good business practice.