From Data Networks to Haptic Networks: Defining the Tactile Internet

The Tactile Internet is formally defined by the International Telecommunication Union (ITU, 2014) as a network combining ultra-low latency with extremely high availability, reliability, and security to enable real-time transmission of haptic information — tactile and kinesthetic sensations — alongside conventional audio-visual data streams. It represents the next evolutionary paradigm of global communications infrastructure, moving beyond the transmission of data, voice, and video to enable real-time haptic interaction, teleoperation, and skill delivery over networks.

The core technical challenge identified across nearly all retrieved sources is the 1 ms round-trip end-to-end latency target — simultaneously a wireless transmission problem, a network architecture problem, and a physical-limit problem constrained by the speed of light over distance. Critically, this constraint cannot be met over distances beyond approximately 150 km via light-speed propagation alone, which is why AI-driven model mediation at the network edge has emerged as the consensus architectural response.

The enabling technology stack comprises five principal sub-domains: Ultra-Reliable Low-Latency Communication (URLLC); Edge/Multi-level Cloud Computing including mobile edge computing (MEC), software-defined networking (SDN), and network function virtualization (NFV); Haptic Codecs and Communication Protocols standardised under IEEE 1918.1.1; Model-Mediated Teleoperation using AI-driven virtual environment proxies at the network edge; and Network Architecture Redesign optimised for the Tactile Internet’s stringent QoS requirements. According to IEEE, the 1918.1 working group established by Delft University of Technology in 2019 remains the only broadly recognised framework for Tactile Internet interoperability.

Seven Years of Research Acceleration: The Innovation Timeline

Tactile Internet research progresses through three discernible phases spanning approximately seven years of active acceleration, from the first directly relevant publications in 2016 through to the most recent vertical-application and B5G/6G-focused work in 2023. Understanding this trajectory is essential for R&D teams benchmarking their entry point against the field’s maturity curve.

The Foundational Phase (2016–2018) is anchored by Beihang University’s two earliest technical papers on URLLC transmission and energy-efficient design for the Tactile Internet (2016), and Ericsson Research’s landmark vision paper on haptic communications over 5G (2017). System architecture concepts including SDN-based intelligent core networks and multi-level cloud models were formalised by Russian institutions during 2017–2018.

The Development and Standardisation Phase (2019–2021) shows the highest density of results. Delft University of Technology described the IEEE 1918.1 standards working group in 2019, marking a transition from research exploration to standardisation. The year 2021 alone accounts for six sources in the dataset, including comprehensive survey articles from Auckland University of Technology, École de Technologie Supérieure (Montréal), and the University of Patras — a signal that the field had reached sufficient maturity for systematic literature review.

The Refinement and Vertical Application Phase (2022–2023) sees focus shift to specific technical optimisations and vertical deployments: deterministic network calculus models for bounding Tactile Internet delays (Hunan Normal University, 2022), fiber-wireless offloading strategies (Beijing Institute of Technology, 2022), and the first B5G/6G-integrated survey (Shri Mata Vaishno Devi University, 2023).

“2021 marks the inflection point: the year comprehensive survey articles from three continents confirmed the Tactile Internet had moved from exploratory research to a field ready for systematic engineering and standardisation.”

Four Technology Clusters Enabling Sub-Millisecond Haptic Interaction

The Tactile Internet research landscape organises into four distinct technology clusters, each addressing a different layer of the latency and reliability challenge. Together they form an integrated stack — from the physical wireless transmission layer up to the application-level haptic interaction model.

Cluster 1: URLLC Physical and MAC Layer Design

This cluster addresses the fundamental wireless transmission layer requirements — short packets, finite blocklength codes, subchannel allocation, and ultra-high reliability targets. The overriding technical constraint is that conventional Shannon-capacity-based design is inadequate for the micro-packet, sub-millisecond regime required by the Tactile Internet. Beihang University’s 2016 work proposes subchannel diversity assignment and bandwidth minimisation under reliability constraints using finite blocklength channel codes. Yonsei University (2019) addresses heterogeneous traffic — kinesthetic, tactile, and high-quality video — with differentiated latency and reliability requirements, proposing novel URLLC techniques for sporadic, medium-to-large packet delivery.

Cluster 2: Network Architecture — Edge Computing, SDN, and Cloud

This cluster addresses the system-level architecture required to achieve the 1 ms latency target. The canonical approach combines mobile edge computing (MEC) to reduce propagation distances, SDN for centralised control-plane intelligence, and NFV for flexible service chaining. Bonch-Bruevich State University of Telecommunication (Russia, 2018) introduces SDN in the cellular core with multi-level MEC, demonstrating round-trip latency significantly below 1 ms in simulation. Futurewei Technologies’ FlexNGIA (2020) proposes a flexible next-generation internet architecture addressing fundamental limitations of current network design for VR, AR, holoportation, and telepresence use cases. Beijing Institute of Technology (2022) formulates minimum-latency resource management as a mixed-integer non-linear problem over three-tier heterogeneous fiber-wireless (FiWi) networks, solved via Lagrange multiplier methods.

Cluster 3: Haptic Communications and Model-Mediated Teleoperation

This cluster covers the haptic data processing pipeline — codecs, protocols, and the model-mediation approach. St. Petersburg State University of Telecommunication (2019) identifies the speed-of-light propagation barrier as the dominant challenge and proposes AI-driven virtual models at edge cloud units to mediate latency over long distances. InterDigital Europe (2021) reviews pre-standardisation activity around bidirectional haptic control, highlighting technology gaps and recommending open research topics, with edge computing emphasised as the key enabler. The Federal University of Rio Grande do Norte (2022) implements a tactile glove master device communicating bidirectionally with a robotic phantom omni manipulator, modelling coarse roughness, fine roughness, and smoothness sensations in a real-time test platform.

Cluster 4: QoS Provisioning, Standardisation, and Performance Bounds

This cluster encompasses the standards framework, QoS measurement methodologies, and formal performance analysis tools necessary to engineer and certify Tactile Internet deployments. Delft University of Technology (2019) describes the IEEE 1918.1 Tactile Internet architecture framework, differentiating it from 5G URLLC and establishing the foundation for the IEEE 1918.1.1 haptic codec sub-standard — the only broadly recognised framework for Tactile Internet interoperability. Hunan Normal University (2022) develops a deterministic network calculus (DNC) analytical model to derive end-to-end delay performance bounds, validating against traffic parameters including arrival rate and burst size. According to ETSI, deterministic performance bounds are a prerequisite for safety-critical network certification.

Application Domains: Where Tactile Internet Meets Real-World Deployment

The Tactile Internet research dataset identifies five primary application domains, with smart manufacturing and healthcare representing the most technically mature verticals for near-term deployment — driven by regulatory and safety standards that define latency tolerance more precisely than consumer perception alone.

Healthcare and Tele-Surgery

Multiple survey papers in the dataset cite tele-surgery and remote medical procedures as the highest-value Tactile Internet application, requiring certified sub-millisecond haptic feedback loops between surgeon and robotic instrument. The Auckland University of Technology survey (2021) explicitly identifies Healthcare 4.0 as a primary Tactile Internet target. The École de Technologie Supérieure comprehensive survey (2021) lists tele-surgery as the canonical latency-critical Tactile Internet use case. Both papers also identify skill-set delivery over networks — enabling remote haptic training for surgery, craft, or mechanical work — as one of the paradigm-shifting societal implications of the Tactile Internet.

Smart Manufacturing and Industry 4.0

Two independent publications from the University of Patras (Greece, 2021) address smart manufacturing as a Tactile Internet application domain, covering human-machine interaction with haptic and tactile sensations, delay mitigation for manufacturing control loops, and the integration of Tactile Internet with augmented and virtual reality for industrial processes. Sejong University’s IoTactileSim (2021) proposes a virtual testbed using Mininet and CoppeliaSim for investigating QoS in tactile Industrial IoT (IIoT), directly operationalising Tactile Internet for industrial robotic teleoperation. Standards bodies including ISO are increasingly engaged in defining performance requirements for haptic-class industrial control systems.

Autonomous and Connected Vehicles

The Islamic University of Madinah (2019) and the Auckland University of Technology survey (2021) both identify autonomous vehicle control and cooperative driving as critical Tactile Internet use cases, where vehicle-to-infrastructure haptic-class latency is required for safety-critical manoeuvre coordination.

Immersive Reality (VR/AR/Haptic Gaming)

Ericsson Research (2017) and Futurewei’s FlexNGIA (2020) both cite virtual reality, telepresence, augmented reality, and holoportation as primary Tactile Internet application domains. Yonsei University (2019) specifically lists immersive VR as a URLLC use case requiring multi-modal haptic plus high-definition video delivery.

Geographic and Institutional Landscape: Who Is Building the Tactile Internet

Among the directly Tactile Internet-relevant results in the dataset (approximately 20 sources), no single commercial entity dominates by publication volume — the landscape is distributed across academic and research institutions globally. Russia and China each contribute three sources and show the earliest and most sustained technical engagement, while Europe drives standardisation and the United States contributes through commercial R&D subsidiaries.

The notable commercial and industrial contributors in the dataset are Ericsson Research (Sweden/global), Futurewei Technologies (US/China), and InterDigital Europe (UK). Standardisation leadership is concentrated at Delft University of Technology (IEEE 1918.1 WG). The geographic white spaces are significant: among the dataset’s directly Tactile Internet-relevant contributors, Brazil and India are present but underrepresented relative to their network infrastructure investment trajectories. These markets represent potential deployment opportunities and partnership targets for Tactile Internet platform providers, particularly in remote healthcare and industrial automation. Research bodies such as OECD have documented the accelerating digital infrastructure investment in both regions.

“No single commercial entity dominates the Tactile Internet research landscape by publication volume — innovation is geographically broad, with Russia and China showing early and sustained technical engagement, and Europe driving standardisation.”

Emerging Directions: B5G, 6G, and Deterministic Performance Guarantees

The most recent publications in the dataset (2022–2023) reveal five forward-looking directions that will define the Tactile Internet’s trajectory toward post-5G implementation and safety-critical deployment certification.

B5G/6G Integration and Reconfigurable Intelligent Surfaces

The 2023 survey from Shri Mata Vaishno Devi University identifies reconfigurable and intelligent wireless infrastructure — a hallmark of 6G — as a prerequisite for real-time Tactile Internet interoperability across heterogeneous networks. This points to the field’s trajectory toward post-5G implementation, where network intelligence is embedded in the physical layer itself. Global 6G standardisation efforts, tracked by ITU under IMT-2030, are expected to formalise these requirements through 2030.

Deterministic Network Performance Guarantees

The 2022 work from Hunan Normal University signals a shift from probabilistic latency guarantees toward deterministic worst-case bounds — a requirement for safety-critical Tactile Internet deployments including surgical robotics and autonomous vehicles. The deterministic network calculus (DNC) model developed validates against traffic parameters including arrival rate and burst size, providing the formal analytical tools needed for regulatory certification of Tactile Internet services.

Fiber-Wireless (FiWi) Hybrid Backhaul Architectures

The 2022 FiWi offloading study from Beijing Institute of Technology points to the emergence of multi-tier heterogeneous wired-wireless backhaul as a practical pathway to meeting Tactile Internet latency budgets beyond pure wireless approaches. The minimum-latency resource management problem is formulated over three-tier heterogeneous FiWi networks, with Lagrange multiplier methods providing tractable solutions to what would otherwise be computationally intractable mixed-integer non-linear optimisation.

Tactile Industrial IoT and Virtual Testbeds

The IoTactileSim work from Sejong University (2021) and the real-time test platform from Federal University of Rio Grande do Norte (2022) indicate growing investment in simulation and emulation infrastructure specifically designed for Tactile Internet, enabling QoS validation before physical deployment in industrial settings. The Federal University of Rio Grande do Norte platform implements a tactile glove master device communicating bidirectionally with a Matlab/Simulink robotic phantom omni manipulator, modelling coarse roughness, fine roughness, and smoothness sensations.

AI-Driven QoS Frameworks for Beyond-5G Tactile Internet

The 2022 Universitat Pompeu Fabra review signals active consolidation of AI-driven, network-slicing-based, and MEC-integrated QoS frameworks as the dominant architectural paradigm for B5G Tactile Internet deployment. Network slicing, mobile edge computing, and AI-driven resource management are identified as the three pillars of QoS provisioning in the Beyond-5G era for Tactile Internet services.