- The Internet of Vehicles has very strict requirements on real-time functionality, security, and connectivity.
- In an IoV cloud, fast-moving scenarios create frequent base station jumps, which not all cloud solutions can accommodate.
Dalian Roiland Technology is a leading solutions provider for Chinese Internet of Vehicles (IoV) and smart vehicles. The company first started constructing an IoV cloud platform eight years ago, since which time it’s morphed into one of the China’s leading IoV cloud platforms with over 2,000 physical servers.
Roiland provides cloud applications, such as car sharing, vehicle control, and smart driving technology on mobile Internet for a number of services. These include new and used auto sales, vehicle repairs and maintenance, insurance, and transportation services.
Growth in the IoV industry has followed a terraced model, with a long fallow period of waiting and knowledge gains followed by an industry-wide surge upwards. When the last surge occurred, Roiland was unable to rapidly expand its IoV services to keep up with demand with the IoV cloud platform it had built in-house. In mid-2015, Roiland started to look at third-party public clouds, but finding one suitable for IoV services wasn’t easy.
IoV levies very particular demands on cloud that not all can fulfill. Describing the extreme requirements of IoV, Roiland CEO Tian Yunong explains that: “You have Internet clouds, IoT clouds, and ubiquitous network clouds that are Internet plus IoT. Roiland’s IoV cloud is a ubiquitous network cloud, which is the next generation of intelligent IoV cloud platforms.” He believes that it’s also the most complex type of cloud application, because it involves people-to-things and things-to-things connections, as well as direct people-to-people connections. “As a result, it involves interaction between people, vehicles, and sensors,” says Tian.
He describes the requirements of IoV, as “extreme,” but what exactly does this mean?
Tian pinpoints five main features:
Always connected: Roiland’s IoV application is a remote-control feedback system that comprises onboard equipment, a cloud platform, and an app. As it involves the safety of vehicles and people, a clear channel of communication between each of these features must be maintained so vehicles can always receive control commands and upload data to the cloud control platform.
Mobile: Traveling at very high speeds can cause delays and interruptions in communication. However, IoV must support connectivity requirements in scenarios of up to 150 kph.
Real time: IoV must capture the data of all collisions involving moving vehicles. As physics holds that a collision takes 20 milliseconds, data collection has to be completed within 10 milliseconds and then transmitted within another 10.
Handling massive amounts of data: With every component in an intelligent car constantly generating data about its operating status, a single vehicle can produce up to 1 GB of data per second. So, the IoV cloud would need to receive the same amount of data in the same timeframe to fully understand the real-time status of a single cloud-controlled vehicle. For an IoV cloud platform, vehicles are like its customers, with each producing data equivalent to at least 10,000 Tencent or Alibaba users. Current communications tech cannot fully support such a large data upload, as a cloud platform would need to support access capabilities for at least 100 Mbps per device.
Security: In IoT, security involves personal possessions and privacy; with IoV it can be a matter of life and death. IoV must be impenetrable.
In July 2016, Huawei and Roiland signed a strategic agreement to develop a next-gen IoV solution and cloud platform. The project, which would also advance the driverless tech and smart vehicle ecosystem, involved connecting Roiland’s IoV cloud platform to HEC to achieve dual-active backup.
How did Huawei meet the rigorous demands of extreme IoV applications? Huawei’s highly flexible and low-latency enterprise-grade Infrastructure as a Service (IaaS) applies its industry-leading FusionSphere cloud operating system, OceanStor enterprise-class storage, FusionServer servers, and software-defined networking and security. It gives Huawei the E2E capability to support enterprise-grade cloud services and user migration to hybrid cloud.
On the service side, HEC stays customer centric. It provides reliable cloud services for industry customers through unique three-level network nodes that cover China’s seven regions and integrate highly available cloud resources.
“Our common background allowed us to come together with Huawei. IoV is a typical ICT application, which meshes precisely with Huawei’s expertise,” states Tian, when talking about Huawei’s superior understanding of IoV application scenarios and related technologies compared with other vendors. “And Huawei was able to provide a complete solution comprising chips, devices, and a platform; for example, SDN, unified IoT gateways, and an open IoT device platform.”
HEC focuses on enterprise-level cloud infrastructure services based on open collaboration, integration and, also in Huawei’s favor according to Tian, its strict commitment to network security and maximizing gains for the customer – Huawei doesn’t monetize customer data.
As Tian points out, Roiland’s IoV cloud platform is different to other clouds. In an IoV cloud, fast-moving scenarios create frequent base station jumps; for example, a vehicle moving at 100 kph switches to a new base station every three seconds. Other clouds are unable or haven’t been designed to support this kind of high-speed mobile scenario. The global service capabilities offered by the HEC and Huawei’s quick response times were perfect ingredients for IoV cloud service provision that could give Roiland a powerful boost in the IoV arena. “Huawei’s research capabilities and adaptability were also particularly strong,“ says Tian. “Huawei was able to get on board very quickly after learning about our particular scenario.”
He admits that at first the project met with some difficulties when integrating Roiland’s IoV platform and HEC. Huawei’s cloud, for example, initially mistook Roiland’s IoV upload feature for second-level data as a DDoS attack, although the feature was an “extreme” yet normal service type. However, Huawei quickly adapted to the scenario, reconstructing the solution and producing new iterations of the platform architecture. Over more than one year of collaboration, Tian has been most impressed by Huawei’s “combat-level response.” HEC’s performance, security, and services have combined to solve all the issues Roiland faced, and no problems have occurred that needed escalating to the vendor’s 800 call center.
Roiland’s IoV cloud platform provides a number of services for downstream customers, including professional IoV cloud services like PaaS and SaaS for IoV enterprises; pre-installation and smart IoV solutions for auto manufacturers; RCTS, UBI, RDM, and RLAS for auto finance and insurance companies; post-installation smart IoV solutions for car sales and service companies; and intelligent urban transportation solutions based on next-gen smart IoV.
Roiland is now mapping out a next-gen smart IoV cloud platform. In collaboration with Huawei, the company hopes to form an industry alliance with upstream and downstream partners to promote the construction and development of the next-gen IoV and a cloud industry ecosystem. Tian outlined a number of expectations for next-gen smart IoV, including a standards system, driverless technology, broader global customer service capabilities, and an ecosystem alliance.
With car exports growing in China – the world’s largest car producer – Roiland hopes to provide IoV services globally in a more efficient way, using Huawei’s global technical service capabilities and cloud data center resources.