OneSoft Solutions: The Future Of O&G Machine Learning Pipeline Integrity
Summary
- OneSoft offers a novel software-as-a-service machine learning service focused on oil and gas pipeline data aggregation and is quickly building an unbreachable moat.
- O&G pipeline companies accrue a net savings of 3.3 million dollars simply by switching legacy processes to OneSoft's software.
- OneSoft currently has no competitors, no debt, and plenty of cash. The global pipeline evaluation business is $1.1 billion and the TAM has potential for growth (water/sewer, and railroads).
- OneSoft's software enables O&G companies to easily meet new regulatory laws.
Editor's note: Seeking Alpha is proud to welcome Soren Rasmussen as a new contributor. It's easy to become a Seeking Alpha contributor and earn money for your best investment ideas. Active contributors also get free access to SA Premium. Click here to find out more »
OneSoft Solutions (OTCQB:OSSIF) offers a novel software-as-a-service (SaaS) machine learning service focused on pipeline data aggregation and intelligent predictive solutions. I believe new contracts with oil and gas (O&G) pipeline companies throughout 2021 will create a "data advantage" for OneSoft. This will create an unbreachable moat centered around OneSoft's one-of-a-kind predictive capabilities, attractive SaaS cost savings, and the easy ability for companies to transition into new regulatory laws.
The ability to reduce unnecessary costs, repairs, and digs is a huge asset to O&G companies. This technology has the opportunity to revolutionize the ~$1 billion industry. Stricter regulations mandate that O&G companies transition legacy processes to smart cloud-based data aggregation. Cognitive Integrity Management (CIM) offers simple compliance tailored to O&G regulations and makes this process seamless. The recent contract with The Fern River Pipeline (Berkshire Hathaway subsidiary - 1 billion dollar pipeline) highlights OneSoft's impending "data advantage." This further leads to "consumption economics." As OneSoft's predictive capabilities improve, O&G companies will demand OneSoft's machine learning capabilities.
Operations
OneSoft Solutions provides a cloud-based, SaaS pipeline integrity management (PIM) suite to the oil & gas industry. OneSoft's mission is to predict pipeline failures and thereby save lives, benefit the environment and reduce O&G companies' costs through machine learning, data science, and cloud computing. OneSoft's PIM system, Cognitive Integrity Management, is powered by Microsoft Azure that utilizes machine-learning and enables pipeline operators to avert pipeline failures and easily comply with regulations. CIM operates through multi-year contracts, ranging from 3-5 years in length, and deploys consumption economics, meaning that as more significant new pieces of intellectual property, functionality, and solutions are added to CIM, OneSoft can increase prices and yield larger returns.
Management and Governance
Insiders own ~34% of OneSoft.
4/5 executives have worked together since 2004.
4/5 board members are independent. All board members own stock.
Executives are compensated by a ~$150-200k salary. Option grants are nominal.
Leadership is heavily invested at ~34% ownership which shows the team is heavily invested and has a large incentive for growth.
Dwayne Kushniruk, the current CEO is scheduled to be replaced by current COO, Brandon Taylor. This was an amicable leadership change and Kushniruk remains on the board and retains all of his shares. In March of 2020, OneSoft started a structured commercialization process (domestic and international) and filled several key positions including hiring a new VP of sales, Technical Sales Manager, and a new Lead Generation Team.
Capital Allocation
OneSoft positioned for cloud computing early and correspondingly exited the on-premise business and returned shareholder capital.
When in need of funds in 2017, rather than issuing common, management sold a portion of its personal stake and used the proceeds to purchase and then exercise warrants, satisfying the company's financing demands while mitigating dilution by 20%.
Customer Value Proposition
"We think that data science machine learning is really the only plausible way that we are going to reduce these pipeline failures." - COO Brandon Taylor
Predictive power can be measured in two ways. The first; the rate of false positives (example: instances in which evaluation calls for excavation, but a failure is not actually imminent). The second; the rate of false negatives (example: imminent failures that are not identified. CIM's deployment reduces false positives by 25-35%. The rate by which CIM reduces false negatives has not been quantified, but the sections below reveal how CIM lessens the probability of pipeline failures.
All customers undertake extensive validations prior to contracting with CIM. Switching costs from legacy processes are material, yet, seven major operators (who collectively represent ~10% of U.S. midstream mileage) concluded that CIM's predictive benefits are sufficient to warrant transition. Of particular note, Phillips 66 became OneSoft's first client when, during trial runs, CIM instrumented the identification of several probable failures that Phillips 66 had overlooked. The benefit was substantive enough that Phillips 66 migrated all of its piggable miles (~55% of total mileage) to a theretofore entirely unproven service for a very sensitive use case.
Other major pipeline operators (Chevron, TransCanada, EnBridge) have attempted (and failed) to build machine learning (ML) software to ingest pipeline data. This suggests that operators believe ML would improve integrity results.
Three major incidents, over the last two decades, have cost over three billion dollars and several hundred deaths. Traditional pipeline evaluation is performed with a "pig" (the name of a pipeline inspection tool). The pig slogs down the pipeline to gather data on the infrastructure of the pipe. Once collected (usually done by a third-party vendor), the data is then uploaded to an Excel file and sent to pipeline operators who tediously comb through vast pages of data; a cumbersome process. Estimates state that only 6% of pipeline data is analyzed. Traditional data collection is slow and often takes up to 6 months to transfer all the data to the pipeline company. This is a slow, manual, and inaccurate process.
Cost Savings
Pipeline operators spend ~$7,700 per mile on integrity management annually, allocated ~$650 to evaluation, ~$1,000 to inspection, and ~$6,050 to maintenance. Estimates conclude CIM currently generates total cost-savings of ~$750 per mile, per discussions with OneSoft management.
Initial signup with OneSoft software offers an average net savings of 3.3 million per company.
CIM's recurring services are priced at ~$45-50 per mile, which is ~90-95% below the cost of legacy data evaluation.
To the extent that CIM reduces the probability of catastrophic failure, it provides material cost-savings: a $1 billion failure costs an operator a sum equal to ~10-15 years of normalized PIM expenses.
CIM reduces excavation costs by lowering the rate of false positives.
Source: SNN Investor Slideshow
Growth Targets
TAM & Moat Focus
The pipeline evaluation segment is a 1 billion dollar per year industry. OneSoft is the sole operator in the pipeline evaluation industry. As OneSoft signs more contracts with pipeline companies and more data is collected, a data advantage becomes imminent.
Source: Sidoti Investor Conference
In the past, CIM's technology only worked with "piggable" pipes (those through which an in-line inspection device can be run). Piggable pipes only account for a small percentage of global pipe mileage. In August 2020, OneSoft announced trialing non-piggable services with a current customer. Non-piggable functionality, which is likely to be fully commercialized during 2021, expands the TAM to substantially all pipeline infrastructure.
Total Addressable Market: What is the total revenue opportunity of ML integrity management solutions?
Pipeline Miles
~4 million global pipeline miles.
~1 million global piggable pipeline miles. Most piggable pipes are transmission (i.e. midstream) miles. This represents OneSoft's current target market.
~3 million distribution miles in the world. Distribution mileage is generally non-piggable and operated by "Local Distribution Companies" (i.e., utilities). OneSoft intends to service distribution miles in the future, however, additional product development is necessary before non-piggable functions can be commercialized.
CIM is pricing its service quite lowly to encourage adoption. Early adoption is particularly critical to CIM, because the data generated by the initial customers are used to improve the service. This leads to better predictive capabilities and further adoption.
For the most accurate view, CIM's pricing should be considered against the cost-savings that it generates for customers. In 2020, recurring revenue per mile will likely be ~$45-50, amounting to just ~5% of the annual cost-savings that CIM provides.
CIM has two pathways through which it can increase revenue per mile. First, as CIM develops more applications, cost-savings will increase because ML will replace and economize inefficient processes; CIM monetizes these developments through modules. Second, in light of the disparity between price and cost-savings, and the switching costs inherent in this service, CIM should have meaningful pricing power over time. In sum, CIM should be able to increase the absolute cost-savings that it provides to customers and capture a greater share of such cost-savings as revenue.
OneSoft management is unsure where revenue per mile may reside in the long-run, however, the company has guided to $100 CAD (~$75 USD) as a near-term target, and COO Brandon Taylor roughly estimates that revenue per mile may increase 2-3x from present levels over the next 5-10 years. Revenue per mile could be as much as an order of magnitude higher at maturity. Some newly developed modules are currently priced at $600-700 per mile and OneSoft is exploring use cases for CIM which cost $30,000 per mile via legacy methods.
Additionally, non-piggable miles - of which none are currently under contract but which represent the majority of the addressable market - are inspected ~5x more frequently than piggable miles and may thereby generate significantly greater revenue per mile.
Water, Sewer & Rail Miles
OneSoft believes that its technology could be applied to water & sewer pipes and railways with little adjustment.
2.5+ million miles of water & sewer pipeline in U.S.
800 thousand railway miles globally.
There are no commercialized ML solutions for water, sewer and rail integrity management.
The aggregate cost of water/sewer failure is greater than that of O&G pipelines. Consequently, CIM could conceivably earn at least as much revenue per mile in water/sewer as in O&G.
Minimal data on railway failure rates
TAM Math: (4 million O&G miles + 2.5 water & sewer miles + 800 thousand railway miles) * ($75-400 at-maturity revenue per mile) = $550 million-$2.9 billion.
Deep Moat
2.7 million piggable pipeline mileage in the U.S.
Deep Moats: If operators adopt ML for integrity management, will CIM capture that adoption?
OneSoft's primary technical achievements are its "ingestion" and "alignment" ML algorithms. The ingestion algorithm enables CIM to automatically consume Excel spreadsheets; the alignment algorithm enables CIM to homogenize the inspection data from dozens of tool vendors populating such spreadsheets. Other entities who have access to considerable amounts of inspection data, such as operators and service companies, align inspection results by hand. Because the manual process is so inefficient, these entities omit 95%+ of their data from the collection. CIM, however, by virtue of its algorithms, consumes all of the data to which it is privy. Furthermore, CIM fields data from multiple operators and multiple tool types, which provides for the most robust dataset.
Valuation
What has to happen to earn a 20%, five-year IRR?
20% five-year IRR requires 2025 ~$100 million equity value.
OneSoft is rightly prioritizing TAM capture over profitability. Accordingly, EV/Sales is the proper valuation metric at this point in OSS' lifecycle, in my view.
BVP Emerging Cloud Index is trading at ~20x revenue. Accounting for OneSoft's small company and illiquidity discounts and the potential for sector-wide multiple compression, I am underwriting to a 2025 10x revenue multiple.
Base case pathways to $10 million in 2025 revenue: 133,000 miles at $75 per mile or 200,000 miles at $50 per mile.
Goal: Mileage Growth
OneSoft has 100,000 piggable miles in trial. Research suggests that an operator has never trial-ed and not signed a contract.
OneSoft is in engagement, but not trial, with a further 300,000 piggable miles.
When non-piggable functionalities are ready for commercialization (likely at some point during 2021), OneSoft will aim to onboard its current customers' non-piggable infrastructure and sign distribution mileage (~3 million miles of TAM).
OneSoft has engaged with a sewer/water customer. Water/sewer represents 2.5 million miles of TAM in U.S.
Revenue per Mile Expansion
Contracting non-piggable miles.
Adoption of higher-priced modules.
If CIM added zero incremental functionality to the platform and simply took prices up to $75 per mile, it would still only be capturing ~10% of its customers' total cost-savings.
Non-piggable commercialization will likely be at some point in 2021. This will open the door for OneSoft to transfer existing clients' non-piggable infrastructure (~3 million miles of TAM).
OneSoft has also engaged with a water/sewer customer. Water/sewer represents 2.5 million miles of TAM in the U.S.
P&L
At-Maturity Economics: How Profitable is it to Supply Machine Learning Integrity Management Solutions? As per CEO Dwayne Kushniruk:
Proforma cash models estimate that EBITDA margins at $10 mm [~$7.5mm USD] revenue would likely be 45% and scale up to ~70% at $50 mm [~$37.5mm USD] revenue. The biggest cost in scale-up is personnel, which only increases a small amount relative to revenue increase, so a good portion of the future revenue is expected to fall to the bottom line.
Revenue: CIM's revenues can be placed into two buckets. First, there are one-time uploads at onboard - CIM charges a fee each time an operator uploads inspection data. Estimates show that uploads are priced at ~$100 per mile (i.e., if an operator signs on for 10,000 miles, CIM will generate $1 million in one-time revenue).
Second are recurring uploads, subscriptions and modules. Pipelines are typically inspected at intervals between one and five years. New inspections are uploaded onto CIM and represent recurring revenue and CIM charges a monthly subscription fee for its core services. Additional modules may be either one-time or recurring depending on the given module's use case. Estimates show that recurring upload, subscription and module fees are currently generating revenue of ~$45-50 per mile per year.
Historical uploads will likely be ~20-30% of revenue in 2020. As CIM's miles under contract grow, onboards make a smaller impact on the top-line. At maturity, sales will be ~100% recurring.
Revenue is insulated from commodity prices (midstream is least exposed to oil/gas cycles and operators conduct integrity management at all times).
Gross Margin: 75-80% today, 90%+ at maturity:
Royalties to Phillips 66 (~$150-250k per year) do not grow with revenues.
Customer onboarding will become more automated. When new onboards slow/cease (i.e., TAM is exhausted), implementation expenses will be immaterial.
Increasing revenue per mile will accrete to gross margin.
EBITDA Margin: ~0% today, ~50-70%+ at maturity:
Employees: Exec comp will leverage dramatically. Sales and software personnel are highly scalable expense items - i.e., salespeople drive recurring revenue, and software engineers to build algorithmic improvements/modules that are redistributed at a minimal cost.
Sales & Marketing: currently attributable to trials and industry conferences, which will moderate as awareness grows and technology is validated. Similar to implementation expenses, S&M will decline substantially as onboards slow.
G&A: OneSoft has always operated remotely. Overhead is minimal and fixed (e.g., public co costs).
Free Cash Flow Margin: EBITDA * (1 - tax rate):
CapEx: de minimis.
NWC: Possibility of sustainable inflows via prepaid contracts.
No debt = no interest.
Risks
OneSoft has an extremely attractive solution to outdated O&G pipeline inspection; however, are companies ready to make the transition?
One risk is that OneSoft is too early in the process of implementation of computer and machine learning. Data science and AI technology are undoubtedly the way of the future, however, are the industry may not yet be at a point where the cost and computational power are strong enough.
After signing 38,000 miles (four companies) in Q1 2019, OneSoft did not contract another client until October 2020. OneSoft prioritized integrating the 2019 companies and did not have the capacity to add new clients for the rest of the year. To address the constraints, OneSoft hired three salespeople (its first dedicated sales employees) and eased onboarding through software improvements. In August 2020, OneSoft noted that its sales pipeline was the most populous in the company's history. This trend is picking up more momentum going into 2021.
Research indicates that CIM is superior to legacy processes, operators have to endure switching costs to separate from their incumbent methods. One risk is that CIM does not receive further adoption. It is important that OneSoft add 10,000-20,000 miles over the next 12 months and thereby validate that CIM has a product-market fit.
How difficult would it be for another company to gather data?
OneSoft will likely have competition in the future as startups enter the market. OneSoft's first-mover advantage is augmented by their predictive AI capabilities that become more accurate as more multi-year contracts are signed. Here is a message from CEO Dwayne Kushniruk responding to a question about other competition to OneSoft:
From a system perspective, this company is one of probably many at various stages of start-up who are trying to get a foothold in the market. Their particular strategy seems to be placing IOT sensors on the pipe to collect data such as pressure, temperature, flow rate, acoustic (and perhaps other data) and relay them to the cloud, then apply ML and AI to manage/monitor/predict conditions on the pipe at those locations and express results in dashboard views.
Inevitably more companies will enter into the pipeline evaluation business. As Brandon Taylor has reiterated, the initial deal with Phillips 66 gave OneSoft a 3-5 year head start on any other competitor.
With this being said, there is the possibility that a competitor could gain a significant contract with a high-caliber company (like Phillips 66). Presently, it seems unlikely that this would happen, as OneSoft is so much further advanced and well equipped because of its sheer data advantage.
Companies Refusing to Drop Legacy Software
Some oil and gas pipeline companies have been slow and even reluctant to drop their outdated pipeline evaluation processes. This would likely slow OneSoft's short-term growth, but in the long run, I do not see this as being a problem. Recent regulatory changes have begun to mandate that companies switch to cloud-based SaaS technology. Further, these companies have potentially billions of dollars to save by making the switch. It is not a question of if they will switch, but when.
COVID-19
COVID-19 has slowed OneSoft's ability to engage with new customers because of travel restrictions. I do not see this being a problem moving forward into 2021.
Conclusion
We believe that our pioneering adoption of cloud computing dating back to 2011, the kickoff as the only O&G participant of Microsoft's first Accelerator for ML and Data Science in 2016, integration of Phillips IP in 2018, and strong validation of our solution by now 10 clients (many who are high profile in the industry) combine to create a unique opportunity for our company. This underscores the reason we decided to raise more capital in 2019 in order to accelerate new product development to increase our competitive moat... Regarding new developments by AI/ML start-ups, we believe these are actually beneficial to our future, because as the industry moves forward with these projects some of the themes with specific uses will be able to be integrated with our CIM platform (think Apple App store concept). (Source: Dwayne Kushnirk Email Conversation)
The momentum is building and OneSoft sits strategically at the forefront as O&G pipeline inspection pivots to machine learning. As new companies sign contracts with OneSoft the faster and more data advantage OneSoft gleans. This leads to better data aggregation, improved intelligent computer algorithms, and cost savings. I predict OneSoft will onboard several more Fortune 500 companies in 2021 and this will solidify OneSoft as the pioneer and revolutionary leader in O&G pipeline machine learning technology.
This article was written by
Analyst’s Disclosure: I am/we are long OSSIF. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Seeking Alpha's Disclosure: Past performance is no guarantee of future results. No recommendation or advice is being given as to whether any investment is suitable for a particular investor. Any views or opinions expressed above may not reflect those of Seeking Alpha as a whole. Seeking Alpha is not a licensed securities dealer, broker or US investment adviser or investment bank. Our analysts are third party authors that include both professional investors and individual investors who may not be licensed or certified by any institute or regulatory body.
Recommended For You
Comments (28)
