Future cellular Internet of Things (IoT) networks are going to be expected to deliver much lower latency and significantly higher reliability. Getting to that point, however, must be a step-by-step approach, said a telco equipment executive at Mobile World Congress Americas earlier this month.\n\u201cDoing one at a time is not so difficult, but doing both at the same time is a challenge,\u201d said Jawad Manssour, head of Networks Portfolio Management at Product Area Networks with equipment maker Ericsson, during a presentation at the conference.\nEricsson is one of the world\u2019s big three principal base station and cellular equipment vendors, along with Huawei and Nokia. Mobile network provider\u00a0Sprint and Ericsson recently announced that they are building a distributed virtualized core IoT network and an IoT operating system.\n\nManssour said one needs to look at IoT evolution in four distinct genres \u2014 Massive IoT, High-Performance IoT, Critical IoT, and Industrial IoT \u2014 to fully appreciate the challenges of IoT overall as use ramps up. Currently 700 million connections are on cellular technology IoT, but that is expected to scale to 3.5 billion by 2023, he said.\nIndustrial IoT and Critical IoT the hardest to implement\nIndustrial IoT and Critical IoT are the newest IoT themes, and Manssour\u00a0said they\u2019re going to be the hardest to implement. They, more than the other genres, most require the ultra-high reliability, always-on availability and extremely low latency.\nAll elements are needed for sharing the tactile, visual, and sound data required during a remote operation, for example.\nA paramedic\u2019s ambulance, for example, where diagnosis, monitoring, or even surgery (which might take place in such vehicles in the future), will be one of the hardest networking spots because of shifting locations, Manssour said.\nThe IoT genre that has the greatest challenge, however, is Industrial IoT, he said.\n\u201cThe most challenging part, looking into the future, is Industrialized IoT,\u201d Manssour said. And it\u2019s \u201cat the very early stage in research.\u201d\nIndustrial IoT is the connected, digitalization of industry overall, and one ripe for short-range 5G use. Standardization is the tricky issue with this genre, he said. Latency thresholds, which he thinks have to be less than 2 milliseconds wirelessly, is one significant area that needs standardization.\n\n\n\n\n\nRobot motion control, assembly line functions, industrial sensors, supply, inventory and delivery management \u2014 all controlled by a networked Operations Control \u2014 are factory applications that 5GACIA, an industry alliance body,\u00a0said are use cases for 5G IoT and would thus fit into Manssour's\u00a0definition of Industrial IoT.\nInterestingly, the Industrial IoT genre will need to be tied into traditional, existing wired networks, 5GACIA said.\n\u201c5G is likely to support various Industrial Ethernet and Time-Sensitive Networking (TSN) features,\u201d 5GACIA says on its website. \u201cThereby enabling it to be integrated easily into the existing wired infrastructure, and in turn enabling applications to exploit the full potential of 5G.\u201d\nMassive IoT and High-Performance IoT implmenetations already underway\nMassive IoT and High-Performance IoT, on the other hand, are both already-existing technologies,\u00a0Manssour explained. Massive IoT is taken care of to a certain extent by cellular links such as NB-IoT and low-power wide-area network (LPWAN).\nMassive IoT includes the low-cost, low-energy sensors, which are expected to be implemented in suitably massive numbers, using low data volumes, all connected to the network. The subset is driven to a certain extent by power requirements \u2014 the batteries mustn\u2019t need to be replaced too often.\nHigh-Performance IoT requires a more challenging increase in throughput and lower latency than Massive IoT. It can include video, like one needs in transportation environments.\nIt\u2019s in early adoption, Manssour said, citing a project Erisson has been working on with the Swiss railway network.\u00a0In that project, mobile broadband users on a train need to share limited bandwidth with high-definition, train monitoring video and allow the train operator to claim priority bandwidth.\nOther High-Performance IoT use cases include camera drones and slotting robots into the network to make robotics cheaper. Network slicing, which involves splitting the network into use-areas, can help in these scenarios, Manssour said.