Chapter 2 - Problem Domain
From Kb
Contact Article Author | Blog of Article Author | FirstPartners.net Home | LinkedIn profile of Author
Contents |
Problem Domain
The Office as a Factory
The great leap forward of the industrial revolution was the assembly line. Raw materials and parts went in one end, labour and machinery was applied to it, with finished products (cars, washing machines, ships) produced at the far end. The assembly line is still present in today’s knowledge economy. Workers may have shirts and ties instead of boiler suits; offices may be air-conditioned and not heated by a blast furnace. There may be more variety in what is produced (an insurance application may have thousands of permutations, rather than the 30 or so colours available in car manufacture) but the assembly line process is similar.
[Ref for 30 or so colours]
In both sorts of assembly line, raw materials (an insurance application, an expense request, a request to trade shares) enter at one end. Various people work on it – in sequence (as a chain of loosely or strictly defined steps) or in parallel. Tools are applied to it (a PC instead of sledgehammer). The end result is some product that while intangible, has added enough value to justify the labour and capital employed.
[sample industries e.g. Insurance Industry , Financial Industry , Pharmaceutical .].
[Picture – office as a production line]
On a traditional assembly line, workers had very little freedom of action [ref]. Unskilled workers could be trained in a matter of minutes to tighten a particular bolt, clip on a car door or spray a particular area. While perhaps personally unsatisfying for the worker , the ’Command and control’ system was easy to administer and supervise. Perhaps fortunately, these approaches do not work with ’knowledge companies’
[ref to Taylor]
Knowledge workers have a far greater leeway in their work, but still must follow some sort of overall process. For example a share trader may have authority to trade millions of Euro’s on behalf of a bank, but will have limits on the positions that he or she may take , and there will be an end-of-day accounting process. While knowledge workers may have greater visibility of the process than their shop-floor counterpart, it is still rare for an individual to know 100% of the processes of the organisation , from their immediate job specification (e.g. Finance) , the process in departments they interact with (e.g. Back office,Sales Targets) or supporting departments (Human resources or procurement policies).
[Ref to financial trading]
The importance of getting this right
Most competing organisations have pretty much the same inputs. Car manufacturers have the same raw inputs (metal and other commodities), similar levels of education in their workforce and at similar prices (the raw inputs are traded globally, manufacturers will relocate to regions where labour is cheaper). However, why has Toyota eclipsed GM as the No 1 manufacturer in the world? R&D levels are similar over the last 19 years , so the answer probably lies with the Toyota Production System, the process by which it builds vehicles and the way that it is continually improving this process. Toyota’s competitive advantage is not it’s cars, but the way in which it builds it’s cars.
[Ref to why intangible is easier to copy]
[Ref to Toyota Story / Toyota Production System & Car production figures]
[Ref to Manufacturers relocating]
[Ref to bit about innovation and R&D spending]
Instead of cars, it is information that flows down the modern company assembly line. Companies that manipulate this information better gain an edge over their competitors , yet information workers cannot be controlled by Taylor’s idea of a foreman with a stopwatch. Enterprises attempting to gain an edge through their superior use of knowledge, both in the use of information and the process that has to be applied to it, face the following problems.
[ref to Taylor stopwatch]
- Formal and Informal knowledge
Things are not always done ’by the book’ , but by rule of thumb. How can we capture this quickly without disrupting to the day to day business ? Are we covered when people leave? Can the business scale if we can only ever have one expert?
[ref – x% of the information in a business is learnt on the job]
- Format of knowledge
How can we express the captured information in a format that gives payback to the business (e.g. Electronic instead of a dusty paper copy, not stuck on some file share, readily usable, instead of being locked away in machine-code)? Is this format easy to update? Can everybody use it from one central location (so that copies do not get ’out of synch’)?
- Management of change
Who can change the information and process? Is this the right balance between being too hard to change (and being stuck in a rut) and being too easy (resulting in chaos). If a change is made , will people know about it and will they take any notice?
- Security and Authorisation
Who should have access to information? Does this access change depending on the context of what the user is doing at the time? Is this access read only or read and update?
- Distribution of information
How do we get information to staff when they need it? How do stop them getting flooded with irrelevant / untimely information?
- Project Based working
People don’t always do the same tasks everyday. People come together to form goal driven teams, and disperse when the objectives have been achieved. How can we easily document these outcomes so that they can be reused both over time and over the organisation?
- New Situations and Judgement
How do express our knowledge in such a way that the organisation does not become sclerotic , and that the knowledge can still be applied in new and changing circumstances?
- Collaboration
No task is done in isolation. How do ensure that tasks and team members collaborate effectively?
- Cross organisation knowledge
In the old days everything was done in house, now the office as a factory must also seamlessly interlink with other suppliers , plugged in as part of the process. How can we implement this?
[Ref to Ford’s massive plant]
[May need more references to tie business costs to these – e.g. Staff turnover costs business X billion dollars per year]
===Traditional Solutions===
Most Enterprise Information (EIS , Enterprise or Business) systems are an attempt to capture and manage knowledge. Most EIS perform 3 broad functions
1. Capture information , e.g. via a web interface (Presentation Layer)
2. Apply business knowledge to this information (Business Layer)
3. Store, forward or display this information (Service or Data Layer)
[3 Tier Diagram]
It is this 2nd,business ,layer that we are most concerned with in this thesis. The presentation and service layers , while not trivial, are known problems that lend themselves to some degree of standardization. In contrast the business layer will be unique to each organisation as it reflects the process and knowledge of the organisation. In some ways, the business layer is the ’learned memory of the organisation’.
[Ref to learned memory / neural networks]
Despite (or perhaps because of) years of implementing EIS systems, many of them suffer from the following problems in the business layer
- All three layers tend to be tightly coupled , so it is not easy to extract the business logic and process contained for use elsewhere
- Business knowledge and rules expressed as code; This is hard to audit and leads to discrepancy between the documentation and the actual implementation.
- It is hard for the domain experts (the guys with the business knowledge) and the technical experts to collaborate as they speak (literally) different languages.
- It can be difficult to update , both in implementation and for fear of un-desirable side-effects .
- While theory states that these functions should be separated, the fact that the business tier is often expressed in programming language like Java means that other functions (e.g. Db access) often creep in over time. Even worse as there is no clearly delineated place to put business logic, it can become scattered throughout the
- Duplication of business knowledge across systems (and become out of synch) can lead to contradictions.
- It is hard for systems to collaborate across processes.
New Solutions
This document concentrates on the following techniques to address some or all of these problems :
- Search
Where you have the information somewhere, but you don’t quite know where it is , or who has it.
- Rules
Where knowledge exists in people’s heads , but needs to be captured in a format that can by easily read by humans and machines.
- Workflow
Where knowledge must be applied step by step as part of a process
- Enterprise Web 2.0
How to distribute and share knowledge in the enterprise using Web 2.0 and Ajax techniques.
This dissertation introduces Red-Piranha, an open source framework for knowledge management. It shows how to use and extend this framework to address information management problems. In particular , we look at how the Red-Piranha framework provides these solutions and integrates with other open solutions to create a modern information assembly line.
Summary
Information and knowledge management is the key that gives companies a competitive edge. Traditional EIS solutions have often failed to deliver on their investment in this regard. This Theses proposes to look at 4 key areas, implemented by the Red-Piranha framework: Search, Rules, Workflow and Enterprise Web 2.0.
Update this structure with the updated one from the .txt file.
This thesis is structured as follows. Chapter 2 examined the problem domain of Knowledge Management. Chapter 3 looks at the Red-Piranha frameworks and details how it can be applied to the Enterprise. It also looks gives a business summary of the samples accompanying this thesis. Chapter 4 describes the functionality and design of the available RP components and suggests how they can be integrated with other open frameworks to extend the frameworks functionality. Chapter 5 gives an in depth look how the 4 off the shelf solutions (Search/Rules/Workflow/Web 2.0) while Chapter 6 gives technical details on how to use these components in an enterprise context. Chapter 7 concludes with a review of the thesis and outlines possible improvements to make to the framework and makes suggestions for future research. A list of references are provided at the end and Appendices.
[Add summary of references]
