?
Hei Huzhi is built
Although we are talking about artificial intelligence, machine study all the time, but reckon a lot of people can feel to not have feeling, without the concept. A lot of people also have doubt: What place is in reflect artificial intelligence, how · of · of · of open artificial intelligence
As it happens sees Forbes an article, say artificial intelligence and machine study. There is Tan Lingren inside dizzy old idea, pare artificial intelligence and machine study reel off raw silk from cocoons however chrysalis, for we are meticulous the most effective method that introduced implementation artificial intelligence wants a place with a few attentions, the viewpoint is original, additional monarch paths, the individual looked to have refresh acknowledge, compile especially arrange existing writings, share everybody.
Hei Huzhi is built
Compile | Old black
Hei Huzhi is built
01
AI and ML
Artificial intelligence (AI) learn with the machine (ML) will tremendous to manufacturing industry generation influence. Had these technologies, manufacturer will obtain stronger computational capability to solve the mankind's inextricability problem. Final aux will be able to is these technologies quite the produces a problem to provide a standard result that production enterprise is exploring all the time since a few centuries. In other words, how do we produce a product efficiently as far as possible, realize 0 waste and least machine down time.
Like the report of most revolutionary technology, right " grail " (compare nowhere the rare world jewellery of look for; The thing that cannot get forever however conceivably hard, here understandable the discussion that is AI and ML) exceeded industry usual practice far. Of AI and ML wish scene the possibility that assumes for us provided a real cause. But, the data infrastructure that because a lot of manufacturer lack win true AI and ML function place,needs, the course that increases perfect production is possible also too abstract, so that let,want to come true its person feels bemused. Does industry leader have doubt via regular meeting " how do we use artificial intelligence technology? Where to use? Where to use??
02
Begin from data
We have a contact to cross many science fiction piece, the AI setting that shows in those films and powerful computational ability, make you fab? It is really, but they also are not completely " eyewash " , before long future, perhaps be in now " scene emersion " .
The AI setting that these cruel dazzle, be by each actual and effective application program " form " , and of these applied processes form criterion only then at data. Actually, data since manufacturer most the asset that did not make full use of, also be the main essential factor with so powerful AI. Consider demand arrangement of Masiluo, a kind by the motive theory of described as pyramid, ground floor is the mainest, the most significant demand, and the demand with top the most complex layer.
Same, mo Ni blocks · Luo Jiadi (Monica Rogati) administrative levels of data science demand also is a pyramid, reflected intelligence of the implementation in producing a system to change requires element. The most rock-bottom demand is, with proper form and system, equal amount, will collect accurate data. The data quality that we collect is higher, the application of artificial intelligence and machine study will be better.
Hei Huzhi is built
03
The quality of data and amount
When beginning to use artificial intelligence, the format memory that a lot of manufacturer discover to their data has a variety of differring is in a few MES, in ERP and SCADA system. If manufacturing process is manual, so the data that collect and analyses is very little, and have a lot of difference. This is so called " dirty data " , the person that this means any trying that understand it - it is data scientist even - will must spend many time and energy. Their need is data changeover general pattern and guide its current system, come with this compose builds a model.
Make sure quality is insufficient merely, data still needs the amount with ensure certain. Once collect good, clean data, production enterprise must ensure they have enough much data, these data are the data that tries to improve process about them, or the data that they try to solve a problem. Production enterprise needs to ensure they have use enoughly exemple, and capture these place is influential using the data of the exemple is variable.
For example, collect the variable of machine Rpm only, still can't tell you malfunctioning reason. But, if you are added,oscillatory, temperature mixes a lot of other bringing about the conditional data of machine error, you can begin to build model and algorithm to forecast breakdown. In addition, as more data collect, you still can establish precision requirement, for example, for example,, This algorithm aux will be able to forecasts this breakdown inside a day quite, accuracy rate is 90 ％ ".
If you feel to collect data very complex, you can use a solution to collect data automatically from all sorts of equipment and system, clear automatically next data or format. Such, engineers are OK and dedicated build model and algorithm at compose, is not to spend time to clear data.
04
Solve a simpler problem above all
The brigade with the methodological open artificial intelligence with first data, allow manufacturer can from the technological process that begins to understand and control them at the beginning. This helps manufacturer control a process neatly not only, eliminate process variable quickly, still can pass more advanced artificial intelligence and machine study model to improve their analytic type.
Remember please: If your process out of control, add artificial intelligence simply and cannot magical rehabilitate it.
Why to want from collect data to solve instant production problem to begin? This still has a crucial reason among them is, it can help you be obtained in the industry send an advantage first. Gu Ge, the company dominant such as Yamaxun and Facebook is worn their industry, because they are the first companies that begin compose to establish data set. Their data set is very giant, data is collected and the analysis is very sophisticated also, this also is the reason that they can strengthen competitive dominant position.
To making likeness also is for the enterprise. Manufacturer begins the brigade of artificial intelligence earlier, they can establish large data set earlier, carry out advanced AI and ML model with this. As every time iteration, the space between they and competitor will be larger and larger.
Using artificial intelligence and machine study is a paragraph itinerary, is not a subtle move that can solve a problem for an instant. It collects data above all simple visible with statistical process in, so that you understand data better,control flow. After this, you will obtain more and more advanced analysis capability, achieve the Utopian goal of perfect production till you, that is to say, artificial intelligence helps you as far as possible efficient, produce a product surely.
