Digital Transformation, Data and Innovation
In my last post I tried to illustrate the importance (and the challenges) of data to digital transformation. This is often a complex and difficult idea for people to understand - why is "data" so hard? Why can't computer systems work more effectively? For example, my father called me over the weekend to ask why his doctors can't get his electronic medical records correct. Trying to explain all of the databases, the types of data, the different sources of the data and the work required to normalize that data is difficult, even to a guy whose job was in computer systems.
Now, if it is difficult to simply compile and use data (like in an electronic medical record) imagine how much more difficult it is to compile data and use it to make decisions. The data has value but often has subtle context associated with it that we humans understand, but machines may not, because they haven't been taught the explicit and implicit meanings of data that accompanies the data (what many people call "meta-data"). For example, a sensor or computer may use computer vision to "see" a school bus, and based on its size and shape may recognize that the vehicle is a school bus.
But the computer may lack the intelligence to also understand the shape and size of the children in the bus (small kids - heading to kindergarten, large kids - heading to high school) or the context (empty bus leaving school, full bus arriving at school). We humans have this insight and understand the less overt concepts around meta-data because we have learned experience. We know that a school bus has a relatively predictable job - taking kids to and from school - and that at different times and different locations the bus is doing a specific job.
What gets even more complicated is when a predictable activity, like a bus trip, takes place outside of normal hours or normal routes. When a school bus is used for a field trip or for a sports event, the bus is outside of its normal routes, operating outside of its normal times. Even we humans may look at a school bus on the highway late in the evening and wonder what they are doing. Machines can only parse through rules or assumptions when data is out of regular bounds.
Data and digital transformation
There is no digital transformation without data. It's as simple as that. Digital transformation is either involved in creating more data (using sensors and IoT devices to gain more data), managing and understanding the data (Big Data, predictive analytics) or using data to make decisions or take actions (autonomous vehicles, robots, AI and machine learning). None of these things can be accomplished without data. Once you realize how important and how valuable data is to digital transformation you'll think again about where and how digital transformation can be successful.
I spoke recently with a software partner to ask them how to help clients prioritize what projects they should start for digital transformation. There were several activities that if they could be automated and digitized would have high return, but obtaining and using the data was proving difficult. His response was to find the problems or challenges where obtaining and using the data was the easiest, not necessarily where there was the biggest return, because access to good, useful data is that important.
Digital Transformation and Innovation
So, what does this tell us about the impending merger of innovation and digital transformation? Again, data can be an input to innovation - helping drive the design of new products or identify market needs and gaps, or it can be the result of innovation, creating new products that generate useful data that can be gathered, analyzed and perhaps monetized.
In both activities - digital transformation and innovation - people, consultants, thought leaders - will try to convince you that the digital tools (AI, machine learning, robotics) or key innovation methods or activities are what matter. Please ignore these arguments. Where innovation and digital transformation are concerned, what matters is the data. Data coming into the activity or data resulting from the activity, and the value you can create from the data heading in either direction.
Now, if it is difficult to simply compile and use data (like in an electronic medical record) imagine how much more difficult it is to compile data and use it to make decisions. The data has value but often has subtle context associated with it that we humans understand, but machines may not, because they haven't been taught the explicit and implicit meanings of data that accompanies the data (what many people call "meta-data"). For example, a sensor or computer may use computer vision to "see" a school bus, and based on its size and shape may recognize that the vehicle is a school bus.
But the computer may lack the intelligence to also understand the shape and size of the children in the bus (small kids - heading to kindergarten, large kids - heading to high school) or the context (empty bus leaving school, full bus arriving at school). We humans have this insight and understand the less overt concepts around meta-data because we have learned experience. We know that a school bus has a relatively predictable job - taking kids to and from school - and that at different times and different locations the bus is doing a specific job.
What gets even more complicated is when a predictable activity, like a bus trip, takes place outside of normal hours or normal routes. When a school bus is used for a field trip or for a sports event, the bus is outside of its normal routes, operating outside of its normal times. Even we humans may look at a school bus on the highway late in the evening and wonder what they are doing. Machines can only parse through rules or assumptions when data is out of regular bounds.
Data and digital transformation
There is no digital transformation without data. It's as simple as that. Digital transformation is either involved in creating more data (using sensors and IoT devices to gain more data), managing and understanding the data (Big Data, predictive analytics) or using data to make decisions or take actions (autonomous vehicles, robots, AI and machine learning). None of these things can be accomplished without data. Once you realize how important and how valuable data is to digital transformation you'll think again about where and how digital transformation can be successful.
I spoke recently with a software partner to ask them how to help clients prioritize what projects they should start for digital transformation. There were several activities that if they could be automated and digitized would have high return, but obtaining and using the data was proving difficult. His response was to find the problems or challenges where obtaining and using the data was the easiest, not necessarily where there was the biggest return, because access to good, useful data is that important.
Digital Transformation and Innovation
So, what does this tell us about the impending merger of innovation and digital transformation? Again, data can be an input to innovation - helping drive the design of new products or identify market needs and gaps, or it can be the result of innovation, creating new products that generate useful data that can be gathered, analyzed and perhaps monetized.
In both activities - digital transformation and innovation - people, consultants, thought leaders - will try to convince you that the digital tools (AI, machine learning, robotics) or key innovation methods or activities are what matter. Please ignore these arguments. Where innovation and digital transformation are concerned, what matters is the data. Data coming into the activity or data resulting from the activity, and the value you can create from the data heading in either direction.
Voted 2nd best innovation blogger two years in a row!
0 Comments:
Post a Comment
<< Home