SmartHelio Autopilot
SmartHelio Autopilot is a physics-informed AI platform that predicts, locates and prices faults across solar PV and battery storage portfolios, then turns them into maintenance tickets, cleaning schedules and reports. It needs no extra hardware, only SCADA access.
What is SmartHelio Autopilot?
SmartHelio Autopilot is the core platform of SmartHelio, a Swiss EPFL spin-off, and it is where the company's performance models, diagnostics, predictions and automation for solar and storage operations actually live. It presents itself as a single source of truth for everyone touching a photovoltaic asset: the operations team, the field technician, the asset manager and the executive who only wants portfolio-level numbers.
The engine is physics-informed AI. Rather than learning fault signatures purely from labelled examples, Autopilot recognises patterns in voltage, current and temperature using models grounded in the physics of PV systems, the product of fifteen years of Swiss research and USD 25 million invested in R&D. SmartHelio claims this cuts the data requirement by roughly 90% compared with conventional AI and, more to the point, keeps the system explainable: each classification comes with the reasoning behind it instead of a black-box score.
Onboarding needs no additional hardware. Autopilot connects through an API to the SCADA or CMS already in place, pulls equipment specifications and reports, and builds a digital twin of the plant in minutes. Output starts on day one, and the vendor's FAQ promises usable results within a week.
Day to day, the Fault Dashboard predicts faults in real time, localises the failing component, pinpoints root causes, quantifies the losses and ranks what to fix first. Diagnostics Theatre lets analysts explore the detected fault signatures and build their own weather or performance benchmarks. The Site Validation Report pushes assignments to ground staff, who act and report back through a technician app. Reports, automatic waterfall loss diagrams and a C-suite portfolio dashboard are produced without manual work, and alerts reach the team over WhatsApp, email or Teams.
Battery storage sits in the same view: co-located PV and BESS monitoring, custom thresholds on temperature, state of charge and voltage, state of health and degradation tracking, and OEM warranty eligibility.
SmartHelio reports 8 GW under management, more than 20,000 plants onboarded, over 5 billion labelled data points and an average 8% performance increase. Named users include Verbund, Schneider Electric, GALP and Daystar Power.
What it does
- Predict faults and component failures on solar plants before they cause downtime
- Locate the failing component inside the plant and identify the root cause
- Quantify energy losses and rank them automatically in a waterfall diagram
- Recommend what to repair first, by severity and by projected loss
- Create, assign and track maintenance tickets through pre-set workflows until resolution
- Generate customisable reports on KPIs, losses and intervention impact without manual work
- Monitor co-located PV and battery storage in one view, from state of charge to OEM warranty status
When to use SmartHelio Autopilot / When not to
A quick filter to help you decide if SmartHelio Autopilot is the right fit.
When to use SmartHelio Autopilot
- Commercial and industrial solar operators running several sites and wanting to know where to send a crew first
- Utility-scale operators and IPPs whose portfolio is growing faster than their O&M headcount
- Solar asset managers who need loss quantification and portfolio-level reporting they can hand to stakeholders
- O&M teams and field technicians, who receive assigned faults with instructions and answer back from a technician app
- Operators of co-located battery storage who want SoC, state of health, degradation and OEM warranty tracking in the same view as the PV plant
When not to use SmartHelio Autopilot
- Homeowners with a residential rooftop array: the product is sold and priced around portfolios, not single installations
- Buyers who expect to see a price, sign up online and start the same afternoon: every route leads to a sales conversation
- Plants with no SCADA or CMS and no API access to their own data, since onboarding depends entirely on that feed
- Teams working on non-solar generation, wind included, which is only touched by SmartHelio's separate climate forecasting product
- Developers looking for a documented product API to build their own tooling on top: none is published
- Organisations whose procurement requires a named hosting jurisdiction, a DPA or a security certification before signature
How to use SmartHelio Autopilot
A typical end-to-end flow, from setup to results.
- Book a demo on the SmartHelio site, or contact the partnerships team directly
- Describe the plant or portfolio and its operational constraints during that first conversation
- Open API access to the existing SCADA or CMS and supply the plant metadata
- Let Autopilot extract equipment specifications and build a digital twin of the plant, a matter of minutes
- Read the first results from day one; consolidated output is promised within a week
- Work the Fault Dashboard daily: predicted faults, their location, root causes and the losses attached to each
- Dig into Diagnostics Theatre to examine fault signatures and set your own benchmarks
- Assign faults through the Site Validation Report; technicians act and send feedback from the technician app
- Receive critical and predicted-fault alerts on WhatsApp, email or Teams
- Use the automatic reports and the C-suite dashboard for portfolio-level review and stakeholder updates
Pros & Cons
Pros
- No additional hardware: the platform plugs into the SCADA or CMS already installed
- Fast onboarding, with a digital twin built in minutes, results from day one and consolidated output within a week
- Explainable by design: the physics-first approach means every fault classification comes with its reasoning, not a black-box score
- Broad compatibility claimed across plant sizes, new and legacy installations, and all major inverter and panel manufacturers
- Measured customer outcomes: Daystar Power reports fault resolution four times faster, a 10% performance uplift and 100% detection accuracy within four months
- Complete operational loop, from detection and prioritisation to ticket, field intervention and report, with battery storage covered alongside PV
- Flexible commercial models, including a free trial, an annual SaaS subscription, a one-off audit and pay-per-fault billing
Cons
- No public pricing at all: three commercial models are described without a single figure, and every path leads to a sales call
- No terms of service and no legal notice on the site; the only legal document is a privacy notice that has not been updated since December 2022
- No hosting jurisdiction disclosed, only a reference to a secure, closed cloud environment
- No security certification published, no data processing agreement and no subprocessor list to hand to a procurement team
- GDPR compliance is never claimed outright, and no Article 27 representative or DPO is designated even though the vendor is established outside the EU
- A genuine technical prerequisite: without API access to a SCADA or CMS, nothing starts
- No documented product API, a single published contact address and no help centre or documented support channel
Pricing & Plans
There is no permanent free plan. A free trial is offered, through the demo booking page rather than a self-service sign-up. No price is published for any of the three commercial models, on the pricing page or anywhere else in the site's source, so the cheapest paid entry point cannot be stated in any currency; quotes are issued on request after a conversation with the sales team.
- annual subscription for monitoring large portfolios
- aimed at utility-scale plants and long-term asset partnerships
- with real-time multi-site monitoring and ongoing diagnostics
- alerts and analytics. Price on request.
- a single detailed audit report for a solar asset
- with loss quantification and root-cause insight
- charged as a fixed fee based on plant size
- with fast turnaround from minimal data and no long-term commitment. Price on request.
- fault-based billing suited to pilot runs and small plants
- where you pay only for what is detected
- with no upfront commitment and conversion to SaaS at any time. Price on request.
Data, GDPR & hosting
A consolidated view of how SmartHelio Autopilot handles your data.
GDPR overview
The privacy notice, generated with Termly and last updated on 15 December 2022, devotes a full section to legal bases and names both the GDPR and the UK GDPR: consent, legitimate interests, legal obligations and vital interests are each set out. Data subjects in the EEA, the UK and Canada are told they may request access, rectification, erasure, restriction and portability, and may complain to their local supervisory authority; Swiss residents are pointed to the FDPIC. A separate CCPA section covers California. What is missing matters as much: SmartHelio never states in so many words that it is GDPR compliant, designates no Article 27 representative and no data protection officer despite being established outside the EU, publishes no data processing agreement and names no subprocessor. All requests go to a single address.
Who owns the data?
There is no terms-and-conditions page anywhere on SmartHelio's site, so ownership of the operational data a customer feeds into Autopilot is not addressed in any public contract. The only legal document published is a privacy notice dated 15 December 2022, and it covers the personal data of visitors and prospects, not plant telemetry. Within that narrower scope, SmartHelio sarl states that it processes no sensitive personal information, receives none from third parties, and has neither disclosed nor sold personal data in the preceding twelve months. On the product side, the single ownership claim made anywhere is that insights rest strictly on the customer's own validated operational data, encrypted in transit and at rest.
Reuse rights
Nothing on the site restricts what a customer may do with what Autopilot produces, and nothing formally grants it either, since no terms of service exist. In practice the platform is built for onward sharing: reports combining KPIs, waterfall loss diagrams and intervention impact are described as tailored for both internal and external use, and insights are meant to be shared with stakeholders at every level. That is a strong indication that outputs are the customer's to circulate, but it remains a product description rather than a licence. On the personal-data side the privacy notice grants access, rectification, erasure and portability rights to people in the EEA, the UK and Canada, exercised through a request form or by email.
Data retention & training
Hosting summary
SmartHelio names no hosting country and no hosting region anywhere on its site. The only architectural statement is that its AI agent operates in the company's secure, closed cloud environment, and the FAQ adds that all data is encrypted both in transit and at rest and that the platform complies with industry standards for data privacy and security, without naming a single standard. The vendor is established in Switzerland, at Prilly, with entities in the United States and in India, so data may plausibly move between jurisdictions, but nothing on the site documents that. The domain itself is served behind Cloudflare, which says nothing about where the application and its data actually sit. No data processing agreement is offered, no subprocessor is disclosed and no security certification is published. For any buyer with a data residency requirement, this is the single largest gap in the vendor's public information and should be settled in writing before onboarding.
Things to keep in mind
Risks and trade-offs to weigh before adopting SmartHelio Autopilot.
- Prediction fatigue: SmartHelio itself insists on human oversight, and a team that stops questioning the dashboard loses exactly the judgement the tool is meant to sharpen
- A missed earth fault has physical consequences; Schneider Electric's own testimonial points to overvoltage and fire risk, so alerting must never be the only safety layer
- No hosting jurisdiction is disclosed, yet plant telemetry is sensitive industrial data that reveals output, availability and commercial performance
- No data processing agreement, no subprocessor list and no ISO 27001 or SOC 2 certification, which leaves an internal auditor with nothing to work from
- No terms of service anywhere, so liability for an analytical error is not allocated by any contract you can read before signing
- Lock-in risk: the digital twin, the fault history and the benchmarks accumulate inside the platform, so reversibility conditions should be negotiated up front
- The AI is described as trained on more than 8 GW of operational data, without the site ever saying what happens to an individual customer's data or how to exclude it
Setup & Integrations
Technical difficulty
Low for the end user, on one condition. Nothing is installed on site and no hardware is added: SmartHelio connects to the SCADA or CMS already running, extracts equipment specifications and builds the digital twin itself, in minutes. The real prerequisite falls on the customer's technical side, namely opening API access to that system and supplying plant metadata, which needs an operations or IT owner rather than an afternoon. There is no self-service route, so every deployment runs through a demo and a guided onboarding.
Deployment
Integrations
Behind SmartHelio Autopilot
Fundraising
Social
Resources
All the official URLs gathered for verification and reference.
Alternatives
Tools that compete with or complement SmartHelio Autopilot.
Frequently asked questions
What exactly is SmartHelio Autopilot?
Do I need to install extra hardware?
How long does onboarding take?
How does it classify faults, and can I trust the result?
Is it compatible with my inverters and plant design?
Does it cover battery storage as well as PV?
What does it cost?
How is my data protected, and where is it hosted?
Who is behind Autopilot?
Is there a minimum age to use the service?
Should you pick SmartHelio Autopilot?
SmartHelio Autopilot is one of the more convincing pieces of software in solar asset management, because it does not stop at detection. It predicts the fault, says where it is, puts a number on what it costs, opens the ticket, tracks the technician's answer and writes the report. The physics-first approach is not marketing dressing either: it is what lets the platform explain a classification instead of asserting it, and what allows it to run on far less data than a purely statistical model would need. Onboarding is genuinely light, provided you can open an API to your SCADA. The customer evidence is specific and checkable, with Verbund, Schneider Electric, GALP and Daystar Power all named, and figures such as four-times-faster fault resolution and a 10% performance uplift attached to a named portfolio rather than floating free.
The weakness is everything around the product. No price is published, so comparison is impossible without a sales call. There are no terms of service at all, and the only legal document is a privacy notice frozen in December 2022 while the product has clearly moved on. No hosting jurisdiction is named, no security certification is published, no data processing agreement is offered and no subprocessor is listed, which is a real obstacle for any buyer with a procurement checklist. GDPR compliance is never claimed in so many words, and no Article 27 representative is designated even though the vendor sits outside the EU.
If you manage a solar or solar-plus-storage portfolio with a working SCADA, Autopilot deserves a demo. Go into it with a written list of questions on hosting, retention, data ownership and exit conditions, because the site will not answer them for you.
- Choosing a selection results in a full page refresh.
- Opens in a new window.