There is a lot of good news for the transportation market coming from LiDAR technology these days. Field productivity is going through the roof! Static scanners with truck mounted tripods have increased productivity from 6 scans per day to over 60. Mobile terrestrial systems are collecting 500,000+ points per second traveling at 45-50 mph. Airborne systems have increased accuracies and densities by orders of magnitude.
So whats holding back large scale market adoption on every transportation infrastructure project? I just had a great conversation with John Krause, State Surveyor for FDOT Florida Department of Transportation about the challenges of adopting a new technology. Here are our two points of view.
From my perspective as a nimble small technology company, I communicated the frustration of trying to push the technology from the bottom up. I recounted demonstrating our TopoLIFT /VZ400 system to consultants. We thought reducing field time by 80% would be a plus. However we were often informed that DOT procurement practices would pretty much reduce their payment by that 80%, as these policies are based on traditional methods requiring more field time. I quickly learned that were inserting this new exciting technology into an existing, complex process that has been standardized over many years. Those standards and processes dont change quickly.
Hearing Johns perspective from the FDOT side was a real education for me. First I have to say that John and FDOT really do get it with respect to these new technology advances. But the challenges associated with adopting new technology, standards and procedures into a really big agency are many and dauntingat least to me.
John summarized some of FDOTs challenges. As the State Surveyor, John and his team carry the responsibility for understanding the workflow from field to finish. This includes understanding the LiDAR technology itself as well as the acquisition, geospatial processing, QA/QC, model extraction and production operations. Another big concern is the IT requirements associated with such massive amounts of data.
John went on to point out that in this difficult economic environment, there is little room for trial and error. For new technologies applied to existing budgeted projects, failure or increased costs are simply not an option. Despite these tight budgets FDOT has used mobile LiDAR on five various scoped projects across the state in the last 18 months. Most projects are now wrapping up as FDOT continues their efforts to become a better educated customer to their consultants.
Of course the issue of how to apply current FDOT procurement standards to LiDAR data is not trivial. Procurement is based on traditional land survey technology processes. The obvious problem is that LiDAR technology turns many of these processes on their head. Field times cut down by 80-90%, back office processing increased by 200-300% and capital investment approaching $1 million. Nothing quite fits at the moment.
John discussed ways in which procurement could accommodate LiDAR technology. FDOT overhead rates arent capped so capital expense could be pulled in, but they must be audited. Technology could just be folded into design-build projects as they are pretty much lump sum. FDOT is even considering more lump sum projects where a certain technology approach is not required; just get the job done and deliver what FDOT needs at the lowest price possiblemy personal favorite.
So this conversation left me both empathetic of FDOTs challenges and optimistic of their direction. Based on some of the activity I have seen in the industry, it also made me realize that as consultants and technology vendors we could do more to help FDOT and other DOTs adopt LiDAR technology. Heres a short list of suggestions:
1) Stop the Point Cloud Drive-bys!! A consultant rips through town with his mobile scanner and drops off a point cloud into the DOTs lap. The DOT isnt prepared to deal with it. As John has mentioned they need deliverables ready to insert into existing design/engineering processes and workflows. Half-finished projectseven if delivered for freearent helping anyone. . (Dont tell me it doesnt happen, as such data often finds its way to Certainty 3D.)
2) Agree on a QA/QC Procedure — Evaluating point cloud data isnt really that difficult if basic principles are followed: a) There should be lineage between the point cloud and a the sealed survey data and b) there should be metrics established for evaluating that lineage. Agree on the implementation of these two principles and make sure the DOT can evaluate the data independently.
3) Identify the Model Extraction Process First Once you know how the topography model will be extracted, placing requirements on the point cloud data becomes much easier.
4) Define your respective responsibilities in the process Understanding every aspect of say mobile LiDAR technology may not be necessary if one is simply buying the data so long as there is a QA/QC procedure in place. The customer should be able to identify errors, but neednt know the cause or how to fix them. Its often very individual system and process specific, so DOTs shouldnt make consultants problems their problems. Just identify the problem and send it back.
I hope this discussion will result in a real appreciation of FDOTs commitment to LiDAR technology adoption. Im sure the story is the same across the country. As vendors and consultants, we are also FDOT partners. So lets look at ourselves in the mirror and do what we can to tighten up our performance wherever possible. That will be a win-win.