Uhuru has SoftBank blessing to push IoT app-dev platform

Japan’s IoT ecosystem is something of a kingdom unto itself, dominated by some pretty entrenched industrial corporations. However, there are quite a few companies supplying these giants with the skills and services needed to enter the IoT fray. Uhuru is one of these, and we spoke to Chief Strategy Officer Masamichi Tanaka, to get a rundown on how the application and development platform came into being.

Founded twelve years ago, Uhuru began as a Salesforce implementation company, and grew to become one of the company’s main partners in Japan – specializing in CRM, marketing, and transformation projects. Around three years ago, a project with Mitsubishi Heavy Industries woke the company up to the potential for IoT application development services – a path that it has explored extensively.

The Mitsubishi project was to connect wind turbines to the internet, in order to remotely monitor them within the Salesforce ecosystem. Tanaka said that this first IoT project was eye opening, and that it led Uhuru to focus more fully on the IoT. Now, around 40% of its work is in the sector.

Thanks to it being funded by SoftBank, as a large investor, as well as the use of Mbed, Uhuru has close ties to ARM. Tanaka said that SoftBank feeds most of its IoT work, through the SoftBank IoT Partner Program, and that Uhuru is heavily involved in Pelion.

Uhuru’s main customers are Japanese multinationals, including Honda, Komatsu, Mitsubishi Heavy Industries, Mitsui, NEC, and Yamaha. Around 98% of its revenue is done domestically, according to Tanaka, although it is involved in lots of US and UK alliances. Uhuru nets around 30 to 40 new IoT customers annually, and it is looking to build its global presence.

We asked about the differences between the Japanese IoT market and those overseas, and Tanaka pointed out that US firms, by and large, are platforms that that are looking to attract customers. This is almost the reverse of Japan, according to Tanaka, where the IoT platforms stem from manufacturers that developed them for their own use and then later looked to expand outwards.

Compounding this is the tendency not to collaborate much, which leads to quite an insular corporate existence, as they try to create their own systems and technology. However, these habits show signs of changing, owing mostly to the failures of IoT proofs-of-concept over the past three years. The big Japanese companies are now changing their behavior, and are reaching out to platforms.

Tanaka said that international offerings do have a good chance to enter the Japanese IoT markets, but stressed that the most important thing is getting access to the device layer. He stressed that customer appetite is still in the early stages, but that it will boom once the R&D cycles kick in. He remains very optimistic for the next few years, especially off the back of big projects, where Uhuru gets to be a valuable part of a larger scheme.

As for threats to this outlook, Tanaka believes that there is still, at the end of a day, a two to three year first mover advantage. Consequently, there’s a big opportunity to grow the IoT concept into something much bigger, and that AI and blockchain technologies could be useful here. However, if a company can’t get on that first-mover bandwagon, it risks being left behind, as the industry moves onwards.

In Uhuru’s view, there are a couple of main factors behind the lack of an IoT boom. The first is that there is too much data for the cloud to handle, in terms of processing capacity and backhaul bandwidth. The second is a lack of engineers, as these proof of concepts require a broad skillset. That lack often leads to the failure of a PoC, which then leads to corporate disenchantment, as it appears that the PoC could not lead to profits.

The core product is Enebular, a development and operations service, which Uhuru licenses to its customers, who then can sell it on to their own customers. Essentially, Enebular lets a customer build an IoT application, and then deploy the application across the necessary devices – an IoT Orchestration Service.

Enebular is compatible with several cloud platforms, including ARM’s Mbed ecosystem, as well as AWS and Heroku, as well as compatible with containerization systems like Docker and Node-RED. A distributed compute architecture is also available, to relieve cloud workloads, for edge-processing. Additionally, there are AI-based tools and visualization systems, which are almost a given these days.

Enebular can be deployed on any IoT layer, according to Tanaka, thanks to its use of JavaScript, so that it can power end-devices, gateways, or even cloud functions. As a development environment, it uses a graphical interface not unlike Node-RED, in which users can build their applications – using the drag-and-drop functions that have proven quite popular among IoT platforms.

Prior to the Mitsubishi deal, the Enebular software was being used as a tool to connect data sets, rather than devices. Back then, Tanaka said that security issues were the source of many problems, and that this led to working closely with ARM – thanks to the Mbed OS, and then later the Pelion IoT platform.

Uhuru emphasizes that this is an answer to the fact that device and cloud development have been largely kept entirely separate, and that Enebular is a way to unify this development environment. To this end, manufacturing, preventative maintenance, asset tracking, and facility management, are all highlighted as ideal use cases – which isn’t surprising, given Uhuru’s customer-base.