by Gwilym Roberts, Chairman, Kilburn & Strode and Costas Stephanides, Technical Manager, RWS IP Services
To launch or not to launch? Patent Landscaping is an essential part of the innovation life-cycle. We are often asked to carry out a “landscaping” exercise for clients and the first thing that we always do is to check exactly what is required. The term “landscaping” is used for a whole range of different approaches, and if we go down the wrong route it can be costly or even impossible to get the relevant information out of the exercise.
There are two typical approaches. The first is called Freedom to Operate, or “FTO”, but which is perhaps better thought of as a risk management exercise. With FTO, the client has a specific product with detailed design information and wants to assess whether it will run into third-party IP problems. The second approach is a broader look at the product in the context of the general development environment. For example, at this stage, a client may have a rough idea that they want to enter a certain product area, but before making design decisions, especially those which may lend themselves to patent protections, they want a better understanding of what is out there already. Tools certainly exist for doing this, but the output needs to be considered very carefully and the reasons for obtaining the information need to be clear.
The focus of this article is patent searching, but it is important to bear in mind that a whole range of rights may need to be considered – brands, copyright elements, trade secrets, registered and unregistered designs and so forth. The whole picture needs to be considered carefully and the question that you need to ask is, “What are the IP related risks of launching this product or process?” After answering that, you must take sensible approaches to limit the scope of the exercise which can otherwise become untenably broad.
1. Separate the wheat from the chaff
The FTO/Risk Management scenario often relates to a fairly complex product (or process), so the first step is to “parse” the product or process into separate elements, focussing on any elements which could meet the technical requirements for patentability. In practice, this could be a very large number of components.
The task can be cut down significantly by first identifying generic features. These are components which are simply standard to the art. Simplistic examples are screws, conventional fastenings, generic computer processors, etc., but it may be that there are multiple elements of the product that have been in the public domain for long enough (20 years) for those elements to be unpatentable. However, care must be taken with this. As is well known, while it is not possible to patent the prior art, it is possible to patent any non-obvious variation on the prior art and this can leave a lot of room for seemingly trivial improvements to actually be the subject of existing third-party IP.
2. Outsource the risk
Once all “generic” features have been excluded from the exercise, the next question is the extent to which components have simply been bought in from suppliers. In those circumstances, it is still wise to consider the positioning of the IP. There must be warranties in place and these need to be checked carefully, but a quick piece of due diligence to assess whether those components have faced problems in other environments should be considered. An indemnity can provide legal cover and damages cover but will not stop an injunction being issued against you; this cannot be transferred. The impact of proceedings being mounted, even with no real prospect of success, needs assessment from a cost, management and adverse publicity perspective. Again, care needs to be taken during this step.
There must be warranties in place and these need to be checked carefully, but a quick piece of due diligence to assess whether those components have faced problems in other environments should be considered.
Hopefully, at the end of this exercise, the actual number of features that need to be considered are significantly reduced and the search can begin in earnest. Typically, a search strategy is created, results filtered and prioritized and, if problems are identified, further work can be done.
3. Define the search
Starting with the search strategy, there are a range of approaches one can take. For example, some clients may be comfortable to focus only on granted patent applications, understanding the risk that pending applications could go to grant and indeed that unpublished applications may not yet have appeared. This is a risky strategy, but it does help manage the extent of the exercise. Territories need to be considered. Infringing acts include making, selling and importing and, as a result, countries in which these acts occur should be targeted. The relevant subject matter can be refined using patent classification codes, by relying on keywords, or indeed a combination of both. The cost can vary widely depending on the strategy adopted and again the risk/benefit profile can inform these budgetary decisions. More limited strategies can include simply focussing on known parties who are likely to launch an infringement action (as opposed to, for example, simply seeking a royalty); in many cases one or two key competitors can form the basis of an exercise. With each of these strategies, it is important to understand the risks of any reductions in scope of the exercise.
Even with these filters in place, a large number of relevant rights can still be uncovered. Bear in mind that in a search of this nature, the search does not focus on the description of the patent but on the claims which define the extent of protection of the patents. As a result, the match is not of an entire complex document, but just a few lines of text against the potential infringement. Additionally, patents and applications can be found in multiple jurisdictions and the legal tests for infringement can vary between those jurisdictions. Tens, hundreds, even thousands of results can be found and additional processing may be required in order to render the exercise manageable.
4. Red – Amber – Green
There is a wide range of options for such management. A “traffic light” system can be implemented where the risk level associated with each document is classified e.g. red/amber/green – the client may decide to only focus on the highest risk cases again with an understanding of the possibilities of omitting relevant documents by taking this approach. More complex presentations of risk matrices can work as well where the balance between the likelihood of the risk and the impact of the risk can be presented. In many ways it is important to establish who the target audience is for the exercise. It could be lawyers, engineers, investors or the Board – the presentation needs to be carefully tailored to those considerations.
Often practitioners are trained to jump straight to the question of whether a specific claim is infringed by a specific product or process element, considering the complex rules of patent claim interpretation. This is a luxury – it is very far down the line that these questions come up. Even when the most relevant patents or applications have been identified and the corresponding potential infringing features compared against them, other issues can include whether the patent is valid, whether there are commercial opportunities such as licensing, cross-licensing or acquisition, whether it would make sense to design-around the relevant IP or indeed it would be a better move to move onto a different project and give up altogether. All of these considerations form sensible, educated business decisions based on the IP risks identified.
At the end of the exercise, if handled correctly and fashioned appropriately, comprehensive risk management can be performed within budget providing appropriate intelligence as to whether to launch the product or process and in what territories.
The second scenario is the “landscaping” scenario where a client wants to map out the opportunities for a specific product or process area. In practice, this is effectively impossible and generally incredibly costly. Firstly, there is a conceptual problem. A landscape effectively demonstrates areas of significant activity and “quiet” areas. Which one is the most important? The busy places? The problem there is that there are so many patents it could be difficult to get through the noise. On the other hand, the quiet places may well represent areas which simply do not make commercial sense. The question can be formulated properly, but careful thought needs to be applied to parse the issues correctly.
Even more of a problem is that it is basically impossible to look at a patent landscape and then identify products that will survive purely off the analysis alone. The patent landscape is an enormously abstracted view of a complex web of overlapping patent claims which cannot be represented in one or two dimensions. This may be something that AI will one day be able to do, but, at the moment, finding a combination of features that do not fall within the scope of a large number of patent claims simply by looking at the patent claims however they are presented is nearly impossible.
However, the process can be used intelligently. A landscape analysis is useful for getting a feel for where the “hotspots” of design are and assessing whether that is where the focus should be, or whether there are technical opportunities outside those hotspots. This can give rise to some broad design decisions, at which stage an initial sketch of the key components and operation of a product or process can be mapped out. That can then be used to kick off a risk management or FTO exercise of the type described above, and the risks managed accordingly. Typically, this is not a single step process.
As the product is developed and, potentially in view of IP found, specific features redesigned, the exercise needs to be repeated to ensure that every time there is a new potentially patentable development, not only is patent protection for the inventor considered, but the risk that somebody else has already obtained relevant IP needs to be taken into account.
Therefore, development and the original search strategy need regular review. Other tweaks should be borne in mind, for example, the fact that there is the 18-month publication “black hole”. Top-up searches need to be done at regular intervals to catch any further IP they may not yet have appeared. The exercise can work, but it needs a close liaison between the inventors, the commercial departments, lawyers and the searchers, to ensure that the exercise is designed correctly at the outset and evolves with the product design process.
Defining the landscape
In terms of the specifics of landscaping, one of the critical factors is creating an accurate master data set from which to extract the relevant information. Even here, there needs to be a time and cost consideration. For the most accurate approach, manual inspection and tagging of the data is the preferred option, but this can be very labour intensive and is generally best reserved for the FTO/risk management exercise. The alternative is to create suitable search strings and use all of the corresponding ‘hits’ to form the master data. A careful selection of search strings is critical and underpins the entire analysis; any conclusions based on poor source data are unlikely to give a true representation of the field.
It should be appreciated that it is generally impossible to create the perfect strategy that will identify all references of interest whilst excluding all outliers. In practice, the master data set should contain most of the pertinent references with minimum noise. The main exception is where the landscape focuses on specific competitors; here it is easier to build a complete data set for a specific portfolio (providing all affiliations can be identified).
In the early days of landscaping, visuals displaying big data were sometimes awkward to comprehend. A reaction to this issue can be a tendency to oversimplify the visuals to create a very large report with seemingly repetitious graphs. This approach may be easier to comprehend on the surface but potentially harder to extract the salient points. The situation has improved due to better software and a better understanding of how we as humans can visually digest large complex material.
A helpful report can tell the narrative through careful consideration of the order and type of charts employed with less emphasis on descriptive text; after all, a picture is worth a thousand words. The necessary conclusions can fill in the gaps that may not be apparent from the data visualizations.
A careful selection of search strings is critical and underpins the entire analysis; any conclusions based on poor source data are unlikely to give a true representation of the field.
The target audience also plays a major factor, particularly in how the results are presented. For example, the complexities of the patent classification and legal event codes should be reserved for the eyes of the IP specialist. Although company financial data will not form a part of the patent bibliography, it can be beneficial to incorporate such data when presenting to the Board. It is also worth remembering that the audience is unlikely to contain a statistician or computer scientist. As such, complex visuals can be dismissed or worse, misinterpreted by the recipient, if poorly presented.
Regarding the narrative, the question that needs to be answered should be known from the outset. A desired conclusion may also be known along with an initial plan on how to stitch the information together. However, it is important to allow the direction of the analysis to be dictated by the data itself and not by any preconceptions. The very nature of statistics means that landscaping projects can be massaged to suit a desired outcome. Nevertheless, this is likely to unravel as the project progresses to the next stages of product development.
A successful exercise, therefore, needs a clear picture at the outset and close cooperation between the stakeholders. The engineers, sales team and product developers need an accurate idea of what information they are after. The in-house legal team needs to understand the risks. The patent attorney and search specialist need to liaise, define the question and present the answer with the results, parameters and limitations clearly identified. The collaboration then continues, through any product redesign or development process, as well as to mitigate any problems that are revealed.
FTO is highly complex and there is, sadly, no such thing as total freedom to operate. A well-designed strategy can, however, permit a business to characterize the IP risks associated with the launch of a new product and ensure that they are correctly shared through the supply chain. Done properly, this complex exercise boils down to a simple assessment: is the launch worth the residual risk?