Application of deep-learning to logistic robots
Among the work in logistics, unloading corrugated cardboard boxes on a pallet (depalletizing) is such heavy work that its automation had been strongly requested. But the conventional robot systems needed detailed information of the boxes such as image data to be registered in advance, and this prevented the adoption of them.
IHI and IHI Logistics & Machinery Corp. worked with Kinema Systems Inc. (U.S. start-up) to successfully develop an deep-learning-based depalletizing system, without the advance registration of image data of cardboard boxes. In addition, based on a lot of data already learned, this system recognizes cardboard boxes with various kind of shapes and printed patterns. It enables flexible operations based on recognition using deep-learning, and this improves its performance.
Robotics change outdoor work sites
At outdoor work sites such as logistics, construction and civil engineering, human workers manually operate various machineries and vehicles. In recent years, these kinds of works require further security and safety at these sites, and at the same time, shortage of labor and skilled operators is getting more serious.
As a tackle on these social challenges, IHI is developing technology to retrofit remote control and autonomy capability onto such machinery and vehicles. It provides a system that can operate multiple machineries and vehicles with few operators from a safe operation room.
Some part of the systems, based on the remote control, has already been put to practical use, and IHI also aims to commercialize the autonomous operation technology earlier by verifying the practicability of such technology at actual work sites.
Research on hydrogen energy produced from renewable energy
For realizing a low carbon society, people expect to effectively use carbon-free hydrogen produced from renewable energy. Although hydrogen does not emit CO2 when used, it is hard to be liquefied and is expensive in transportation and storage. IHI focuses on ammonia, whose transportation cost is relatively low as a hydrogen carrier, and aims to create a value chain from production to utilization.
We have developed a technology to use ammonia as a fuel for power production in order to reduce CO2 emission. Because ammonia is difficult to be burned, burner design to co-fire ammonia directly with conventional fuels is the key technological point to achieve stable operation. This time, we have succeeded to co-fire ammonia with pulverized coal or natural gas in our test facilities for coal fired boiler and gas turbine. The realization of a low carbon society is certainly approaching.
Ocean current power generation system: new renewable energy technology using the Kuroshio Current
In July 2017, IHI and the New Energy and Industrial Technology Development Organization (NEDO) have completed a 100kW class prototype subsea floating type ocean current power generation system “Kairyu,” which is a new renewable energy technology that utilizes the energy in ocean currents. In August, “Kairyu” was installed into the waters of the Kuroshio Current off the coast of Kuchinoshima, Kagoshima Prefecture and the demonstration test was conducted. Consequently, the world’s first subsea floating type ocean current power generation system succeeded to generate power from the Kuroshio Current.
The demonstration test proved that “Kairyu” can display the performance as designed and that the subsea floating type ocean current power generation system can be operated and can generate power in the waters. We could acquire a lot of data and know-how for the future practical application.
The IHI group is working on achieving a “smart factory” adapted to enhance factory productivity by fusing manufacturing technology and ICT together. The aims of this are: (1) to quickly grasp disturbance hampering production such as equipment failure to quickly take countermeasures against it; and (2) to improve output per person and equipment/energy utilization efficiency. For example, in a machine factory, by attaching various sensors to processing equipment to analyze data, the signs of failure can be detected in advance. Also, by attaching IC tags, etc. to various components, processing time and waiting time in each processing step, and who is in charge can be grasped without taking time and effort. Analyzing the resulting data allows us to work on identifying processing steps problematic for improving productivity and on effectively improving them.
Data analysis technology for utilizing ICT and IoT
The IHI group is advancing the sophistication of our products and services and manufacturing processes by collecting a lot of operating information on products delivered to our customers and on in-house equipment using our shared platform for utilizing ICT and IoT, called ILIPS (IHI group Lifecycle Partner System). We are developing failure sign sensing technology flexibly responding to changes in operating environment such as seasonal variations and a technology allowing anyone to utilize experts’ knowhow using AI/machine learning on the basis of text information such as maintenance records. We are further working on creating customer value, such as planning optimum maintenance/operation and proposing new business.