Accelerating Satellite Data Collection with Regulatory Reform
This blog has focused extensively on how emerging machine learning techniques can be used against remote sensing imagery to derive insights such as automated building footprint detection and car localization. We believe this type of research is important to revealing how public and private sector organizations could leverage the coming influx of remote sensing imagery from space and unmanned aerial system startups. In fact, we partnered with DigitalGlobe (DG) and NVIDIA to release labeled, very high resolution satellite imagery on Amazon Web Services as part of SpaceNetTM in order to accelerate this exploration. Other organizations such Planet, Draper, Dstl, IEEE and IARPA have all launched competitions focusing on the analysis of satellite imagery. Given the amount of open source research and startup activity in the machine learning field, it is worth taking a moment to examine an important factor influencing how startups and incumbents are able to collect and sell satellite data: regulation.
Wernher von Braun has been famously quoted for saying that “we can lick gravity, but sometimes the paperwork is overwhelming.” Although he was referencing the early tribulations associated with the U.S. manned spaceflight program, the reflection is equally attributable to the current state of the U.S. remote sensing industry. On the one hand, the industry is experiencing unparalleled activity in the commercial markets. Venture capital funding and new company starts are at an all-time high (see CB Insights report) and incumbent aerospace firms have made a series of new product announcements (e.g. DG’s Legion story). On the other hand, there has been a growing list of public and private sector stakeholders bemoaning the current state of the U.S.’s outdated remote sensing regulatory framework (e.g., U.S. Rep. Babin, U.S. Rep. Kilmer, Planet, and DG). While most generally agree that the existing laws need revision and that industry should play a greater role, it remains unclear what steps should be taken in the near term. Before considering some of the proposed legislative solutions, let’s first review the existing framework and the current critiques leveled against it.
3 Agencies. 3 Licenses. 1 Satellite.
Companies seeking to build, launch and maintain a remote sensing satellite constellation must receive regulatory approval from three different agencies: the National Oceanic and Atmospheric Administration (NOAA), the Federal Communications Commission (FCC), and the Federal Aviation Administration (FAA). A majority of the focus from industry stakeholders and lawmakers has been on revising how NOAA’s Office of Commercial Remote Sensing Regulatory Affairs (CRSRA) reviews and issues operating licenses to remote sensing companies, but the FCC and FAA licenses are also essential to securing spectrum rights and launch permission respectively. In addition to revising the existing, individual regulations, it is important to note that the licensing processes conducted by NOAA, FCC, and FAA are largely independent of each other which can create coordination issues within the U.S. Government (USG).
Operational Licensing: Timeliness, Predictability and Transparency
All commercial space remote imaging companies are required to have an operating license issued by NOAA. The original licensing process which dates back to the 1992 Land Remote Sensing Policy Act was designed for companies developing a limited number of large satellites. The development timeline for these systems was usually long and required extensive financing. As one might expect, this licensing process has been unable to keep pace with the new, more rapid development and deployment schedules posed by space startups and incumbents with new product lines. There are three major concerns with the current process: (1) timeliness of review; (2) predictability; and (3) transparency. First, many business leaders have expressed their frustration with the slow decision timelines of the licensing process. Under the current regulations, NOAA CRSRA is supposed to issue a licensing decision no later than 120 days after submission. Some have described this timeline as “almost an impossible feat” for a variety of reasons ranging from staffing issues to the “stop clock” requirement when questions are issued from the USG including national security organizations.
Second, industry’s frustration with licensing delays has been further exacerbated by the lack of predictability of those timelines. Startups and large publicly traded companies alike rely upon predictable processes to project capital expenditures, employment needs and product development timelines, and unexpected licensing delays can have can serious business ramifications. For example, a startup developing its initial constellation would be highly impacted by a prolonged delay because the satellites are its primary, near-term revenue generator as well as the justification for additional venture capital funding.
Third and finally, industry stakeholders have repeatedly expressed their concern with the lack of transparency regarding licensing questions or changes to the existing license conditions. Although the USG has worked to expand its outreach to industry partners, several companies have stated that industry should have a more active role in representing their commercial capabilities and interests. Specifically, industry would like to know what the real concerns are when questions are raised as a result of a license request or during a proposed policy change by organizations such as NOAA CRSRA or the Office of the Secretary of Defense (OSD). For instance, this topic was recently discuss at NOAA’s Advisory Committee on Commercial Remote Sensing (ACCRES) meeting held on April 12th (See meeting minutes). Two commercial representatives asked if the commercial company in question could be present in the room during the interagency review meeting in order to provide business and technical insight. Although the OSD Space Policy representative stated that companies could not be present at interagency review meetings, NOAA reiterated that the Office of Space Commerce is present at the meetings and is there to serve as an advocate for the commercial company.
Spectrum Allocation: Identifying Interference
Remote sensing companies are also required to secure a license from the FCC to use a particular band of spectrum for transmitting and receiving data as well as to submit their orbital parameters plan and mitigation of orbital debris plan. Although companies agree with the purpose of the regulations, they have also expressed some frustration with the increasingly frequent delays. There are two primary causes for these delays: (1) spectrum interference claims and (2) orbital interference claims.
Most companies requesting licensing for a new satellite or constellation are likely to encounter delays due to concerns regarding spectrum interference filed by existing constellation operators. Spectrum interference is one of the leading issues in the satellite communication (“satcom”) industry, especially for providers with satellites located in geostationary orbit (GEO). Despite the legitimate concerns regarding access to spectrum and interference and momentary service disruptions, many interference claims lack detailed analyses quantifying the level of projected risk especially from a low earth orbit (LEO) constellation. As a result, companies must spend time and resources negotiating these claims and quantifying the risk. This activity is particularly challenging for a startup that may have limited resources to engage in prolonged negotiations with a more established provider.
Companies must also work with the FCC to identify and address potential orbital interference issues as part of their orbital parameter and debris mitigation plans. While NOAA and other national security stakeholders provide oversight for orbital interference issues involving USG assets, potential interference issues involving commercially owned and operated assets fall under the purview of the FCC. The recent increase in constellation proposals and orbital complexity has increased the number of petitions filed from commercial companies citing concerns about potential orbital interference issues. The filing of such petitions requires the FCC to mediate these complicated negotiations which entails a nuanced understanding of orbital dynamics and potential operational risks. These complex tasks can create strain on the FCC’s workforce given the number of applications and petitions submitted and under review.
Finding a Middle Path for Regulations
Proposed policy solutions to the existing regulatory framework remain in their formative stages. Stakeholder opinion ranges from advocating for adjustments to the existing system to a “fundamental rethinking” of the licensing process. The most recently circulated draft bill, the American Space Commerce Free Enterprise Act of 2017, proposes several significant changes to the existing remote sensing regulatory framework. According to the analysis of the draft bill, a majority of the regulatory authority would be consolidated under the Department of Commerce Office of Space Commerce (OSC). The overall intent of the bill appears to be to streamline the regulatory approval process. For example, the bill provides automatic licensing approval if the USG cannot provide reasonable questions or concerns within 60 days, which is a major divergence from the existing process.
The goal of any proposed policy changes should be to support the continued development of the commercial sector while maintaining security and stability in the space domain. This general goal has been reiterated in several leading documents since the beginning of the decade: the 2010 National Space Policy of the USA; the 2011 National Security Space Strategy (Unclassified Summary); and the National Geospatial-Intelligence Agency’s (NGA) 2015 Commercial GEOINT Strategy. While it is unclear if the currently proposed legislation will garner sufficient political support, there are several less dramatic but still impactful policy recommendations that stakeholders across the spectrum are likely to support. The list below is not intended to be an all-encompassing alternative to any legislative proposals. Instead, it merely seeks to identify several quick wins that could be incorporated into the existing regulatory framework or adopted into draft legislative proposals.
§ Auto-Approval Determinations: Implementing some form of auto-approval for remote sensing licenses could be very helpful in accelerating review times and easing workforce burden. A set of commonly agreed upon criteria would be used for auto-approval determinations. In fact, members of the ACCRES committee have begun to discuss the adoption of some type of auto-approval function for certain licensing requests (See April 12th minutes). Although the comprehensive criteria would need to be established, a leading variable for electro-optical systems is likely to be resolution. For example, systems with low to moderate ground sample distances (GSD) should receive an automatic approval or abbreviated review timeline.
§ Review Transparency: It is essential that licensees and USG stakeholders including members of the national security community continue to build working relationships. While ACCRES serves as one forum for industry and government organizations to discuss regulatory practices, companies should have additional opportunities to discuss their technical capabilities or address questions with government regulators. Ideally, the USG interagency review committee(s) would be required meet with existing and prospective commercial licensees prior to passing a ruling that impacts the business operations of those companies.
§ Adherence to Timeline: Strict adherence to a set of timelines is essential for remote sensing companies. The current draft legislation calls for a 60-day review cycle which is a significant reduction from the existing policy of 120 days. While the shortened timeline would likely be appreciated by industry, it is arguably more important in the near term that whatever timeline is set is consistently enforced. This means the timeline should not stop or be put on hold while specific questions are being addressed. Instead, stakeholders would have a set amount of time to respond to questions. If the respondents require more time to respond to a question, then they could submit a request for an extension for a set amount of time. This extension could be constituted as stopping the clock, but it would require both government and commercial stakeholders to be responsive to initial questions and requests.
§ Government Coordination: The current process requires commercial companies to submit licensing requests to NOAA, FCC and FAA. The ability to file a single application that would automatically initiate a submission to all three agencies would save companies time and money as well as limit the complexity for new market entrants.
